Skip to content

Reimplement llama with fused MLIR operators + simplify operators#76

Open
andrej wants to merge 121 commits intodevelfrom
alt-simplifying-refactor
Open

Reimplement llama with fused MLIR operators + simplify operators#76
andrej wants to merge 121 commits intodevelfrom
alt-simplifying-refactor

Conversation

@andrej
Copy link
Collaborator

@andrej andrej commented Feb 16, 2026

Operator simplification by @hunhoffe

Added

Changed

Removed

PR Merge Checklist

  1. The PR is rebased on the latest devel commit and pointing to devel.
  2. Your PR has been reviewed and approved.
  3. All checks are passing.

@github-actions
Copy link
Contributor

📊 Test Results for Test Example Applications

04c0147 (2026_03_12_22_55_00)

IRONCLAD

Tested on 2026_03_12_22_55_00 at commit 04c0147.

Test Checks TTFT (mean)TPS (mean)
llama_3.2_1b_prompt_1024_tokens_40 ❌ 0/1 n/a n/a
📈 Trends (vs main branch) for Test Example Applications

04c0147 (2026_03_12_22_55_00)

IRONCLAD Trends

llama_3.2_1b

Commit/Date Num Tokens (max)Num Tokens (mean)Num Tokens (median)Num Tokens (min)Num Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)Total (max)Total (mean)Total (median)Total (min)Total (stddev)
130b6ea — 2025-12-05 21:33:1240.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.71 (-0.42%)4.64 (-0.09%)4.64 (+0.65%)4.55 (-0.22%)0.05 (-17.66%)4.41 (-0.34%)4.39 (-0.19%)4.38 (-0.33%)4.37 (-0.15%)0.01 (-25.90%)12.96 (-0.00%)12.80 (+0.07%)12.80 (-0.23%)12.67 (+0.44%)0.09 (-21.12%)
0a6c11c — 2025-12-03 23:35:1540.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.73 (n/a)4.64 (n/a)4.61 (n/a)4.56 (n/a)0.06 (n/a)4.42 (n/a)4.40 (n/a)4.40 (n/a)4.37 (n/a)0.02 (n/a)12.96 (n/a)12.79 (n/a)12.83 (n/a)12.62 (n/a)0.12 (n/a)

llama_3.2_1b_prompt_1024_tokens_40

No metrics available.

llama_3.2_1b_prompt_13_tokens_1

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:071.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.62 (-4.45%)0.62 (-4.45%)0.62 (-4.45%)0.62 (-4.45%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:231.00 (n/a)1.00 (n/a)1.00 (n/a)1.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.65 (n/a)0.65 (n/a)0.65 (n/a)0.65 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_13_tokens_40

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:0740.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.30 (+2.38%)4.30 (+2.38%)4.30 (+2.38%)4.30 (+2.38%)0.00 (n/a)0.61 (-2.49%)0.61 (-2.49%)0.61 (-2.49%)0.61 (-2.49%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:2340.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.20 (n/a)4.20 (n/a)4.20 (n/a)4.20 (n/a)0.00 (n/a)0.63 (n/a)0.63 (n/a)0.63 (n/a)0.63 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_2048_tokens_1

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:071.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:231.00 (n/a)1.00 (n/a)1.00 (n/a)1.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_2048_tokens_40

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:0740.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)0.00 (n/a)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:2340.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)0.00 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)0.00 (n/a)

1 similar comment
@github-actions
Copy link
Contributor

📊 Test Results for Test Example Applications

04c0147 (2026_03_12_22_55_00)

IRONCLAD

Tested on 2026_03_12_22_55_00 at commit 04c0147.

Test Checks TTFT (mean)TPS (mean)
llama_3.2_1b_prompt_1024_tokens_40 ❌ 0/1 n/a n/a
📈 Trends (vs main branch) for Test Example Applications

04c0147 (2026_03_12_22_55_00)

IRONCLAD Trends

llama_3.2_1b

Commit/Date Num Tokens (max)Num Tokens (mean)Num Tokens (median)Num Tokens (min)Num Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)Total (max)Total (mean)Total (median)Total (min)Total (stddev)
130b6ea — 2025-12-05 21:33:1240.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.71 (-0.42%)4.64 (-0.09%)4.64 (+0.65%)4.55 (-0.22%)0.05 (-17.66%)4.41 (-0.34%)4.39 (-0.19%)4.38 (-0.33%)4.37 (-0.15%)0.01 (-25.90%)12.96 (-0.00%)12.80 (+0.07%)12.80 (-0.23%)12.67 (+0.44%)0.09 (-21.12%)
0a6c11c — 2025-12-03 23:35:1540.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.73 (n/a)4.64 (n/a)4.61 (n/a)4.56 (n/a)0.06 (n/a)4.42 (n/a)4.40 (n/a)4.40 (n/a)4.37 (n/a)0.02 (n/a)12.96 (n/a)12.79 (n/a)12.83 (n/a)12.62 (n/a)0.12 (n/a)

llama_3.2_1b_prompt_1024_tokens_40

No metrics available.

llama_3.2_1b_prompt_13_tokens_1

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:071.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.62 (-4.45%)0.62 (-4.45%)0.62 (-4.45%)0.62 (-4.45%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:231.00 (n/a)1.00 (n/a)1.00 (n/a)1.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.65 (n/a)0.65 (n/a)0.65 (n/a)0.65 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_13_tokens_40

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:0740.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.30 (+2.38%)4.30 (+2.38%)4.30 (+2.38%)4.30 (+2.38%)0.00 (n/a)0.61 (-2.49%)0.61 (-2.49%)0.61 (-2.49%)0.61 (-2.49%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:2340.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.20 (n/a)4.20 (n/a)4.20 (n/a)4.20 (n/a)0.00 (n/a)0.63 (n/a)0.63 (n/a)0.63 (n/a)0.63 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_2048_tokens_1

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:071.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:231.00 (n/a)1.00 (n/a)1.00 (n/a)1.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_2048_tokens_40

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:0740.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)0.00 (n/a)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:2340.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)0.00 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)0.00 (n/a)

@github-actions
Copy link
Contributor

📊 Test Results for Test Example Applications

04c0147 (2026_03_12_23_32_55)

IRONCLAD

Tested on 2026_03_12_23_32_55 at commit 04c0147.

Test Checks TTFT (mean)TPS (mean)
llama_3.2_1b_prompt_1024_tokens_40 ❌ 0/1 n/a n/a
📈 Trends (vs main branch) for Test Example Applications

04c0147 (2026_03_12_23_32_55)

IRONCLAD Trends

llama_3.2_1b

Commit/Date Num Tokens (max)Num Tokens (mean)Num Tokens (median)Num Tokens (min)Num Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)Total (max)Total (mean)Total (median)Total (min)Total (stddev)
130b6ea — 2025-12-05 21:33:1240.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.71 (-0.42%)4.64 (-0.09%)4.64 (+0.65%)4.55 (-0.22%)0.05 (-17.66%)4.41 (-0.34%)4.39 (-0.19%)4.38 (-0.33%)4.37 (-0.15%)0.01 (-25.90%)12.96 (-0.00%)12.80 (+0.07%)12.80 (-0.23%)12.67 (+0.44%)0.09 (-21.12%)
0a6c11c — 2025-12-03 23:35:1540.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.73 (n/a)4.64 (n/a)4.61 (n/a)4.56 (n/a)0.06 (n/a)4.42 (n/a)4.40 (n/a)4.40 (n/a)4.37 (n/a)0.02 (n/a)12.96 (n/a)12.79 (n/a)12.83 (n/a)12.62 (n/a)0.12 (n/a)

llama_3.2_1b_prompt_1024_tokens_40

No metrics available.

llama_3.2_1b_prompt_13_tokens_1

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:071.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.62 (-4.45%)0.62 (-4.45%)0.62 (-4.45%)0.62 (-4.45%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:231.00 (n/a)1.00 (n/a)1.00 (n/a)1.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.65 (n/a)0.65 (n/a)0.65 (n/a)0.65 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_13_tokens_40

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:0740.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.30 (+2.38%)4.30 (+2.38%)4.30 (+2.38%)4.30 (+2.38%)0.00 (n/a)0.61 (-2.49%)0.61 (-2.49%)0.61 (-2.49%)0.61 (-2.49%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:2340.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.20 (n/a)4.20 (n/a)4.20 (n/a)4.20 (n/a)0.00 (n/a)0.63 (n/a)0.63 (n/a)0.63 (n/a)0.63 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_2048_tokens_1

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:071.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:231.00 (n/a)1.00 (n/a)1.00 (n/a)1.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_2048_tokens_40

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:0740.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)0.00 (n/a)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:2340.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)0.00 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)0.00 (n/a)

@github-actions
Copy link
Contributor

📊 Test Results for Test Example Applications

aa82d40 (2026_03_13_22_49_20)

IRONCLAD

Tested on 2026_03_13_22_49_20 at commit aa82d40.

Test Checks TTFT (mean)TPS (mean)
llama_3.2_1b_prompt_1024_tokens_40 ❌ 0/1 n/a n/a
📈 Trends (vs main branch) for Test Example Applications

aa82d40 (2026_03_13_22_49_20)

IRONCLAD Trends

llama_3.2_1b

Commit/Date Num Tokens (max)Num Tokens (mean)Num Tokens (median)Num Tokens (min)Num Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)Total (max)Total (mean)Total (median)Total (min)Total (stddev)
130b6ea — 2025-12-05 21:33:1240.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.71 (-0.42%)4.64 (-0.09%)4.64 (+0.65%)4.55 (-0.22%)0.05 (-17.66%)4.41 (-0.34%)4.39 (-0.19%)4.38 (-0.33%)4.37 (-0.15%)0.01 (-25.90%)12.96 (-0.00%)12.80 (+0.07%)12.80 (-0.23%)12.67 (+0.44%)0.09 (-21.12%)
0a6c11c — 2025-12-03 23:35:1540.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.73 (n/a)4.64 (n/a)4.61 (n/a)4.56 (n/a)0.06 (n/a)4.42 (n/a)4.40 (n/a)4.40 (n/a)4.37 (n/a)0.02 (n/a)12.96 (n/a)12.79 (n/a)12.83 (n/a)12.62 (n/a)0.12 (n/a)

llama_3.2_1b_prompt_1024_tokens_40

No metrics available.

llama_3.2_1b_prompt_13_tokens_1

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:071.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.62 (-4.45%)0.62 (-4.45%)0.62 (-4.45%)0.62 (-4.45%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:231.00 (n/a)1.00 (n/a)1.00 (n/a)1.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.65 (n/a)0.65 (n/a)0.65 (n/a)0.65 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_13_tokens_40

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:0740.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.30 (+2.38%)4.30 (+2.38%)4.30 (+2.38%)4.30 (+2.38%)0.00 (n/a)0.61 (-2.49%)0.61 (-2.49%)0.61 (-2.49%)0.61 (-2.49%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:2340.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.20 (n/a)4.20 (n/a)4.20 (n/a)4.20 (n/a)0.00 (n/a)0.63 (n/a)0.63 (n/a)0.63 (n/a)0.63 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_2048_tokens_1

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:071.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:231.00 (n/a)1.00 (n/a)1.00 (n/a)1.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_2048_tokens_40

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:0740.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)0.00 (n/a)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:2340.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)0.00 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)0.00 (n/a)

@github-actions
Copy link
Contributor

📊 Test Results for Test Example Applications

402eba9 (2026_03_13_23_16_23)

IRONCLAD

Tested on 2026_03_13_23_16_23 at commit 402eba9.

Test Checks TTFT (mean)TPS (mean)
llama_3.2_1b_prompt_1024_tokens_40 ❌ 0/1 n/a n/a
📈 Trends (vs main branch) for Test Example Applications

402eba9 (2026_03_13_23_16_23)

IRONCLAD Trends

llama_3.2_1b

Commit/Date Num Tokens (max)Num Tokens (mean)Num Tokens (median)Num Tokens (min)Num Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)Total (max)Total (mean)Total (median)Total (min)Total (stddev)
130b6ea — 2025-12-05 21:33:1240.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.71 (-0.42%)4.64 (-0.09%)4.64 (+0.65%)4.55 (-0.22%)0.05 (-17.66%)4.41 (-0.34%)4.39 (-0.19%)4.38 (-0.33%)4.37 (-0.15%)0.01 (-25.90%)12.96 (-0.00%)12.80 (+0.07%)12.80 (-0.23%)12.67 (+0.44%)0.09 (-21.12%)
0a6c11c — 2025-12-03 23:35:1540.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.73 (n/a)4.64 (n/a)4.61 (n/a)4.56 (n/a)0.06 (n/a)4.42 (n/a)4.40 (n/a)4.40 (n/a)4.37 (n/a)0.02 (n/a)12.96 (n/a)12.79 (n/a)12.83 (n/a)12.62 (n/a)0.12 (n/a)

llama_3.2_1b_prompt_1024_tokens_40

No metrics available.

llama_3.2_1b_prompt_13_tokens_1

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:071.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.62 (-4.45%)0.62 (-4.45%)0.62 (-4.45%)0.62 (-4.45%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:231.00 (n/a)1.00 (n/a)1.00 (n/a)1.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.65 (n/a)0.65 (n/a)0.65 (n/a)0.65 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_13_tokens_40

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:0740.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.30 (+2.38%)4.30 (+2.38%)4.30 (+2.38%)4.30 (+2.38%)0.00 (n/a)0.61 (-2.49%)0.61 (-2.49%)0.61 (-2.49%)0.61 (-2.49%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:2340.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.20 (n/a)4.20 (n/a)4.20 (n/a)4.20 (n/a)0.00 (n/a)0.63 (n/a)0.63 (n/a)0.63 (n/a)0.63 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_2048_tokens_1

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:071.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:231.00 (n/a)1.00 (n/a)1.00 (n/a)1.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_2048_tokens_40

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:0740.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)0.00 (n/a)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:2340.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)0.00 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)0.00 (n/a)

@github-actions
Copy link
Contributor

📊 Test Results for Small Benchmark/Test Suite

197d469 (2026_03_18_17_02_42)

IRONCLAD

Tested on 2026_03_18_17_02_42 at commit 197d469.

Test Checks Latency (mean)Bandwidth (mean)GFLOP/s (mean)
axpy_1_cols_2_channels_2048_tile_2048_3.0 ❌ 0/5 n/a n/a n/a
axpy_2_cols_2_channels_2048_tile_1024_3.0 ❌ 0/5 n/a n/a n/a
axpy_4_cols_2_channels_2048_tile_512_3.0 ❌ 0/5 n/a n/a n/a
axpy_8_cols_2_channels_2048_tile_256_3.0 ❌ 0/5 n/a n/a n/a
dequant_1_cols_1_channels_2048_tile_2048 ❌ 0/5 n/a n/a n/a
dequant_1_cols_2_channels_2048_tile_1024 ❌ 0/5 n/a n/a n/a
dequant_2_cols_1_channels_2048_tile_1024 ❌ 0/5 n/a n/a n/a
dequant_2_cols_2_channels_2048_tile_512 ❌ 0/5 n/a n/a n/a
dequant_4_cols_1_channels_2048_tile_512 ❌ 0/5 n/a n/a n/a
dequant_4_cols_2_channels_2048_tile_256 ❌ 0/5 n/a n/a n/a
dequant_8_cols_1_channels_2048_tile_256 ❌ 0/5 n/a n/a n/a
dequant_8_cols_2_channels_2048_tile_128 ❌ 0/5 n/a n/a n/a
eltwise_add_1_cols_2_channels_2048_tile_2048 ❌ 0/5 n/a n/a n/a
eltwise_add_2_cols_2_channels_2048_tile_1024 ❌ 0/5 n/a n/a n/a
eltwise_add_4_cols_2_channels_2048_tile_512 ❌ 0/5 n/a n/a n/a
eltwise_add_8_cols_2_channels_2048_tile_256 ❌ 0/5 n/a n/a n/a
eltwise_mul_1_cols_2_channels_2048_tile_2048 ❌ 0/5 n/a n/a n/a
eltwise_mul_2_cols_2_channels_2048_tile_1024 ❌ 0/5 n/a n/a n/a
eltwise_mul_4_cols_2_channels_2048_tile_512 ❌ 0/5 n/a n/a n/a
eltwise_mul_8_cols_2_channels_2048_tile_256 ❌ 0/5 n/a n/a n/a
gelu_1_cols_1_channels_2048_tile_2048 ❌ 0/5 n/a n/a n/a
gelu_1_cols_2_channels_2048_tile_1024 ❌ 0/5 n/a n/a n/a
gelu_2_cols_1_channels_2048_tile_1024 ❌ 0/5 n/a n/a n/a
gelu_2_cols_2_channels_2048_tile_512 ❌ 0/5 n/a n/a n/a
gelu_4_cols_1_channels_2048_tile_512 ❌ 0/5 n/a n/a n/a
gelu_4_cols_2_channels_2048_tile_256 ❌ 0/5 n/a n/a n/a
gelu_8_cols_1_channels_2048_tile_256 ❌ 0/5 n/a n/a n/a
gelu_8_cols_2_channels_2048_tile_128 ❌ 0/5 n/a n/a n/a
gemm_1792x896x1152_64x32x48_8cols_ccolmaj ❌ 0/5 n/a n/a n/a
gemm_192x384x64_48x96x16_4cols ❌ 0/5 n/a n/a n/a
gemm_192x384x64_48x96x16_4cols_bcolmaj_ccolmaj ❌ 0/5 n/a n/a n/a
gemm_2048x2048x2048_64x64x64_1cols ❌ 0/5 n/a n/a n/a
gemm_2048x2048x2048_64x64x64_2cols_bcolmaj ❌ 0/5 n/a n/a n/a
gemm_2048x2048x2048_64x64x64_8cols_bcolmaj_ccolmaj ❌ 0/5 n/a n/a n/a
gemm_384x1536x1792_32x48x64_4cols_bcolmaj ❌ 0/5 n/a n/a n/a
gemm_896x1792x640_32x64x80_8cols_ccolmaj ❌ 0/5 n/a n/a n/a
layer_norm_1_cols_1_channels_2048_tile_2048 ❌ 0/5 n/a n/a n/a
layer_norm_1_cols_2_channels_2048_tile_1024 ❌ 0/5 n/a n/a n/a
layer_norm_2_cols_1_channels_2048_tile_1024 ❌ 0/5 n/a n/a n/a
layer_norm_2_cols_2_channels_2048_tile_512 ❌ 0/5 n/a n/a n/a
layer_norm_4_cols_1_channels_2048_tile_512 ❌ 0/5 n/a n/a n/a
layer_norm_4_cols_2_channels_2048_tile_256 ❌ 0/5 n/a n/a n/a
layer_norm_8_cols_1_channels_2048_tile_256 ❌ 0/5 n/a n/a n/a
layer_norm_8_cols_2_channels_2048_tile_128 ❌ 0/5 n/a n/a n/a
matrix_vector_mul_128x128_32tsi_128tso_1col ❌ 0/5 n/a n/a n/a
matrix_vector_mul_2048x8192_1tsi_1024tso_2col ❌ 0/5 n/a n/a n/a
matrix_vector_mul_2048x8192_1tsi_2048tso_1col ❌ 0/5 n/a n/a n/a
matrix_vector_mul_2048x8192_1tsi_256tso_8col ❌ 0/5 n/a n/a n/a
matrix_vector_mul_2048x8192_1tsi_512tso_4col ❌ 0/5 n/a n/a n/a
matrix_vector_mul_8192x2048_4tsi_1024tso_1col ❌ 0/5 n/a n/a n/a
matrix_vector_mul_8192x2048_4tsi_1024tso_2col ❌ 0/5 n/a n/a n/a
matrix_vector_mul_8192x2048_4tsi_1024tso_4col ❌ 0/5 n/a n/a n/a
matrix_vector_mul_8192x2048_4tsi_1024tso_8col ❌ 0/5 n/a n/a n/a
mem_copy_16_cores_2_chans_2048_tile_128_False ❌ 0/5 n/a n/a n/a
mem_copy_1_cores_1_chans_2048_tile_2048_False ❌ 0/5 n/a n/a n/a
mem_copy_2_cores_1_chans_2048_tile_1024_False ❌ 0/5 n/a n/a n/a
mem_copy_2_cores_2_chans_2048_tile_1024_False ❌ 0/5 n/a n/a n/a
mem_copy_4_cores_1_chans_2048_tile_512_False ❌ 0/5 n/a n/a n/a
mem_copy_4_cores_2_chans_2048_tile_512_False ❌ 0/5 n/a n/a n/a
mem_copy_8_cores_1_chans_2048_tile_256_False ❌ 0/5 n/a n/a n/a
mem_copy_8_cores_2_chans_2048_tile_256_False ❌ 0/5 n/a n/a n/a
mha ❌ 0/5 n/a n/a n/a
relu_1_cols_1_channels_2048_tile_2048 ❌ 0/5 n/a n/a n/a
relu_2_cols_1_channels_2048_tile_1024 ❌ 0/5 n/a n/a n/a
relu_4_cols_1_channels_2048_tile_512 ❌ 0/5 n/a n/a n/a
relu_8_cols_1_channels_2048_tile_256 ❌ 0/5 n/a n/a n/a
rms_norm_1_cols_1_channels_2048_tile_2048 ❌ 0/5 n/a n/a n/a
rms_norm_1_cols_2_channels_2048_tile_1024 ❌ 0/5 n/a n/a n/a
rms_norm_2_cols_1_channels_2048_tile_1024 ❌ 0/5 n/a n/a n/a
rms_norm_2_cols_2_channels_2048_tile_512 ❌ 0/5 n/a n/a n/a
rms_norm_4_cols_1_channels_2048_tile_512 ❌ 0/5 n/a n/a n/a
rms_norm_4_cols_2_channels_2048_tile_256 ❌ 0/5 n/a n/a n/a
rms_norm_8_cols_1_channels_2048_tile_256 ❌ 0/5 n/a n/a n/a
rms_norm_8_cols_2_channels_2048_tile_128 ❌ 0/5 n/a n/a n/a
rope_1c_32rows_512cols_32arows_0m ❌ 0/5 n/a n/a n/a
rope_1c_32rows_512cols_8arows_0m ❌ 0/5 n/a n/a n/a
rope_2c_32rows_512cols_32arows_0m ❌ 0/5 n/a n/a n/a
rope_2c_32rows_512cols_8arows_0m ❌ 0/5 n/a n/a n/a
rope_8c_32rows_512cols_32arows_0m ❌ 0/5 n/a n/a n/a
rope_8c_32rows_512cols_8arows_0m ❌ 0/5 n/a n/a n/a
sigmoid_1_cols_1_channels_2048_tile_2048 ❌ 0/5 n/a n/a n/a
sigmoid_2_cols_1_channels_2048_tile_1024 ❌ 0/5 n/a n/a n/a
sigmoid_4_cols_1_channels_2048_tile_512 ❌ 0/5 n/a n/a n/a
sigmoid_8_cols_1_channels_2048_tile_256 ❌ 0/5 n/a n/a n/a
silu_1_cols_1_channels_2048_tile_2048 ❌ 0/5 n/a n/a n/a
silu_2_cols_1_channels_2048_tile_1024 ❌ 0/5 n/a n/a n/a
silu_4_cols_1_channels_2048_tile_512 ❌ 0/5 n/a n/a n/a
silu_8_cols_1_channels_2048_tile_256 ❌ 0/5 n/a n/a n/a
softmax_2_cols_2_channels_32768_tile_1024 ❌ 0/5 n/a n/a n/a
softmax_2_cols_2_channels_32768_tile_512 ❌ 0/5 n/a n/a n/a
softmax_4_cols_4_channels_32768_tile_2048 ❌ 0/5 n/a n/a n/a
swiglu_decode_1x2048x2048 ❌ 0/5 n/a n/a n/a
swiglu_prefill_256x2048x2048 ❌ 0/5 n/a n/a n/a
tanh_1_cols_1_channels_2048_tile_2048 ❌ 0/5 n/a n/a n/a
tanh_2_cols_1_channels_2048_tile_1024 ❌ 0/5 n/a n/a n/a
tanh_4_cols_1_channels_2048_tile_512 ❌ 0/5 n/a n/a n/a
tanh_8_cols_1_channels_2048_tile_256 ❌ 0/5 n/a n/a n/a
transpose_2048_M_64_N_1_cols_1_channels_64_m_64_n_8_s ❌ 0/5 n/a n/a n/a
transpose_2048_M_64_N_1_cols_2_channels_64_m_64_n_8_s ❌ 0/5 n/a n/a n/a
weighted_rms_norm_1_cols_2_channels_2048_weights_2048 ❌ 0/5 n/a n/a n/a
weighted_rms_norm_2_cols_2_channels_2048_weights_1024 ❌ 0/5 n/a n/a n/a
weighted_rms_norm_4_cols_2_channels_2048_weights_512 ❌ 0/5 n/a n/a n/a
weighted_rms_norm_8_cols_2_channels_2048_weights_256 ❌ 0/5 n/a n/a n/a
📈 Trends (vs main branch) for Small Benchmark/Test Suite

197d469 (2026_03_18_17_02_42)

IRONCLAD Trends

axpy_1_cols_2_channels_2048_tile_2048_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.13 (n/a)0.11 (n/a)0.10 (n/a)0.08 (n/a)0.02 (n/a)162.40 (n/a)120.96 (n/a)117.00 (n/a)94.40 (n/a)23.11 (n/a)

axpy_1_cols_2_channels_2048_tile_2048_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.16 (-3.31%)0.14 (-1.29%)0.15 (+16.18%)0.10 (-6.80%)0.03 (+3.89%)125.70 (+7.25%)93.56 (+2.12%)83.50 (-13.92%)75.00 (+3.45%)22.38 (+18.09%)
84d3478 — 2026-02-17 23:25:130.17 (n/a)0.14 (n/a)0.13 (n/a)0.10 (n/a)0.03 (n/a)117.20 (n/a)91.62 (n/a)97.00 (n/a)72.50 (n/a)18.95 (n/a)

axpy_2_cols_2_channels_2048_tile_1024_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.11 (n/a)0.09 (n/a)0.09 (n/a)0.08 (n/a)0.01 (n/a)152.00 (n/a)135.46 (n/a)142.80 (n/a)115.00 (n/a)15.03 (n/a)

axpy_2_cols_2_channels_2048_tile_1024_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (-0.73%)0.11 (-7.43%)0.11 (-8.63%)0.07 (-27.71%)0.02 (+50.39%)182.70 (+38.30%)120.86 (+11.97%)110.80 (+9.38%)96.10 (+0.73%)35.16 (+122.88%)
84d3478 — 2026-02-17 23:25:130.13 (n/a)0.12 (n/a)0.12 (n/a)0.09 (n/a)0.02 (n/a)132.10 (n/a)107.94 (n/a)101.30 (n/a)95.40 (n/a)15.78 (n/a)

axpy_4_cols_2_channels_2048_tile_512_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.12 (n/a)0.10 (n/a)0.10 (n/a)0.07 (n/a)0.02 (n/a)166.70 (n/a)128.78 (n/a)121.80 (n/a)103.90 (n/a)22.95 (n/a)

axpy_4_cols_2_channels_2048_tile_512_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (-10.91%)0.11 (-7.68%)0.12 (-0.73%)0.06 (-17.94%)0.03 (+0.99%)194.10 (+21.85%)124.52 (+10.29%)103.30 (+0.68%)97.20 (+12.24%)40.32 (+39.15%)
84d3478 — 2026-02-17 23:25:130.14 (n/a)0.11 (n/a)0.12 (n/a)0.08 (n/a)0.03 (n/a)159.30 (n/a)112.90 (n/a)102.60 (n/a)86.60 (n/a)28.97 (n/a)

axpy_8_cols_2_channels_2048_tile_256_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.10 (n/a)0.08 (n/a)0.09 (n/a)0.07 (n/a)0.01 (n/a)184.40 (n/a)153.18 (n/a)144.10 (n/a)127.10 (n/a)21.26 (n/a)

axpy_8_cols_2_channels_2048_tile_256_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (+0.01%)0.11 (+6.10%)0.11 (+0.22%)0.08 (-0.33%)0.02 (-9.83%)156.30 (+0.32%)115.26 (-6.40%)112.70 (-0.18%)95.50 (+0.00%)24.34 (-10.22%)
84d3478 — 2026-02-17 23:25:130.13 (n/a)0.10 (n/a)0.11 (n/a)0.08 (n/a)0.02 (n/a)155.80 (n/a)123.14 (n/a)112.90 (n/a)95.50 (n/a)27.11 (n/a)

dequant_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.00 (n/a)160.40 (n/a)141.12 (n/a)136.40 (n/a)114.30 (n/a)17.37 (n/a)

dequant_1_cols_1_channels_2048_tile_2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-13.00%)0.04 (-0.87%)0.04 (+1.16%)0.03 (-3.17%)0.01 (-31.33%)158.40 (+3.26%)128.12 (-0.19%)123.80 (-1.12%)110.40 (+14.88%)18.36 (-15.65%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)153.40 (n/a)128.36 (n/a)125.20 (n/a)96.10 (n/a)21.77 (n/a)

dequant_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.06 (n/a)0.05 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)133.60 (n/a)106.08 (n/a)95.10 (n/a)87.90 (n/a)17.90 (n/a)

dequant_1_cols_2_channels_2048_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (-3.74%)0.05 (-12.63%)0.05 (-5.21%)0.03 (-33.47%)0.01 (+52.60%)162.70 (+50.37%)110.84 (+18.98%)98.80 (+5.44%)81.50 (+3.95%)31.02 (+149.23%)
84d3478 — 2026-02-17 23:25:130.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)108.20 (n/a)93.16 (n/a)93.70 (n/a)78.40 (n/a)12.45 (n/a)

dequant_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.07 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)143.40 (n/a)101.02 (n/a)97.60 (n/a)78.90 (n/a)22.56 (n/a)

dequant_2_cols_1_channels_2048_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (+8.90%)0.05 (+1.31%)0.05 (-1.85%)0.03 (+6.35%)0.01 (+7.11%)152.30 (-5.93%)115.08 (-1.20%)100.80 (+1.92%)88.10 (-8.23%)27.20 (-4.42%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)161.90 (n/a)116.48 (n/a)98.90 (n/a)96.00 (n/a)28.46 (n/a)

dequant_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.05 (n/a)0.04 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)159.70 (n/a)126.32 (n/a)109.00 (n/a)106.80 (n/a)22.48 (n/a)

dequant_2_cols_2_channels_2048_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.07 (+27.96%)0.05 (+1.57%)0.04 (-18.06%)0.03 (+10.23%)0.01 (+22.79%)164.70 (-9.31%)119.12 (-1.39%)120.20 (+22.03%)75.60 (-21.82%)31.61 (-13.80%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)181.60 (n/a)120.80 (n/a)98.50 (n/a)96.70 (n/a)36.67 (n/a)

dequant_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.06 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)156.10 (n/a)120.86 (n/a)114.90 (n/a)88.30 (n/a)24.40 (n/a)

dequant_4_cols_1_channels_2048_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-2.55%)0.04 (-12.79%)0.05 (-16.25%)0.03 (-12.85%)0.01 (-1.20%)164.60 (+14.78%)122.14 (+15.14%)115.90 (+19.48%)97.50 (+2.63%)25.08 (+19.84%)
84d3478 — 2026-02-17 23:25:130.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)143.40 (n/a)106.08 (n/a)97.00 (n/a)95.00 (n/a)20.93 (n/a)

dequant_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)168.10 (n/a)135.74 (n/a)132.20 (n/a)107.40 (n/a)22.91 (n/a)

dequant_4_cols_2_channels_2048_tile_256_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (+20.83%)0.05 (+8.21%)0.04 (+1.42%)0.03 (+17.39%)0.01 (+14.84%)166.40 (-14.80%)122.34 (-7.97%)127.70 (-1.39%)84.60 (-17.22%)30.47 (-19.21%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)195.30 (n/a)132.94 (n/a)129.50 (n/a)102.20 (n/a)37.71 (n/a)

dequant_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.04 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.00 (n/a)170.40 (n/a)140.12 (n/a)135.80 (n/a)119.60 (n/a)17.94 (n/a)

dequant_8_cols_1_channels_2048_tile_256_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-25.19%)0.04 (-21.02%)0.04 (-19.70%)0.04 (-19.76%)0.00 (-37.92%)142.70 (+24.63%)126.44 (+26.24%)125.70 (+24.46%)115.50 (+33.68%)10.61 (+3.97%)
84d3478 — 2026-02-17 23:25:130.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)114.50 (n/a)100.16 (n/a)101.00 (n/a)86.40 (n/a)10.20 (n/a)

dequant_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.05 (n/a)0.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)198.20 (n/a)160.00 (n/a)166.20 (n/a)106.50 (n/a)29.86 (n/a)

dequant_8_cols_2_channels_2048_tile_128_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.04 (-26.69%)0.04 (-11.93%)0.04 (-16.99%)0.03 (+27.28%)0.01 (-62.60%)193.10 (-21.41%)147.68 (+2.96%)136.10 (+20.44%)135.40 (+36.49%)25.41 (-58.96%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.05 (n/a)0.02 (n/a)0.01 (n/a)245.70 (n/a)143.44 (n/a)113.00 (n/a)99.20 (n/a)61.90 (n/a)

eltwise_add_1_cols_2_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.16 (n/a)0.13 (n/a)0.15 (n/a)0.08 (n/a)0.03 (n/a)145.70 (n/a)99.28 (n/a)83.40 (n/a)77.60 (n/a)28.70 (n/a)

eltwise_add_2_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.12 (n/a)0.10 (n/a)0.12 (n/a)0.08 (n/a)0.02 (n/a)158.10 (n/a)122.52 (n/a)105.00 (n/a)101.40 (n/a)27.53 (n/a)

eltwise_add_4_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.15 (n/a)0.12 (n/a)0.14 (n/a)0.06 (n/a)0.04 (n/a)198.70 (n/a)115.74 (n/a)86.50 (n/a)79.30 (n/a)50.70 (n/a)

eltwise_add_8_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.10 (n/a)0.11 (n/a)0.07 (n/a)0.02 (n/a)176.10 (n/a)125.14 (n/a)112.70 (n/a)111.40 (n/a)28.50 (n/a)

eltwise_mul_1_cols_2_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.15 (n/a)0.13 (n/a)0.14 (n/a)0.07 (n/a)0.03 (n/a)166.10 (n/a)102.68 (n/a)89.20 (n/a)83.70 (n/a)35.57 (n/a)

eltwise_mul_2_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.15 (n/a)0.13 (n/a)0.15 (n/a)0.09 (n/a)0.03 (n/a)134.40 (n/a)96.08 (n/a)82.40 (n/a)81.50 (n/a)22.84 (n/a)

eltwise_mul_4_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.12 (n/a)0.10 (n/a)0.11 (n/a)0.06 (n/a)0.02 (n/a)196.40 (n/a)128.60 (n/a)115.40 (n/a)105.90 (n/a)38.33 (n/a)

eltwise_mul_8_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.10 (n/a)0.11 (n/a)0.08 (n/a)0.01 (n/a)155.20 (n/a)123.34 (n/a)113.70 (n/a)112.20 (n/a)18.35 (n/a)

gelu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.09 (n/a)0.06 (n/a)0.02 (n/a)134.90 (n/a)94.88 (n/a)87.90 (n/a)77.60 (n/a)23.30 (n/a)

gelu_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.10 (n/a)0.11 (n/a)0.07 (n/a)0.02 (n/a)111.40 (n/a)84.60 (n/a)77.20 (n/a)72.70 (n/a)15.66 (n/a)

gelu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.10 (n/a)0.05 (n/a)0.02 (n/a)155.90 (n/a)95.10 (n/a)79.40 (n/a)76.70 (n/a)34.18 (n/a)

gelu_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.10 (n/a)0.11 (n/a)0.05 (n/a)0.02 (n/a)151.70 (n/a)92.64 (n/a)76.60 (n/a)74.90 (n/a)33.24 (n/a)

gelu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.10 (n/a)0.05 (n/a)0.02 (n/a)151.30 (n/a)95.16 (n/a)83.80 (n/a)74.70 (n/a)31.97 (n/a)

gelu_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.09 (n/a)0.10 (n/a)0.05 (n/a)0.02 (n/a)181.50 (n/a)104.40 (n/a)84.90 (n/a)82.90 (n/a)43.16 (n/a)

gelu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.09 (n/a)0.09 (n/a)0.06 (n/a)0.02 (n/a)143.30 (n/a)99.78 (n/a)91.30 (n/a)81.60 (n/a)24.96 (n/a)

gelu_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.07 (n/a)0.08 (n/a)0.05 (n/a)0.01 (n/a)173.70 (n/a)116.88 (n/a)106.20 (n/a)98.20 (n/a)32.02 (n/a)

gemm_1792x896x1152_64x32x48_8cols_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:255.00 (n/a)4.88 (n/a)4.97 (n/a)4.51 (n/a)0.21 (n/a)2083.60 (n/a)1928.60 (n/a)1893.70 (n/a)1882.30 (n/a)86.77 (n/a)1965.38 (n/a)1921.12 (n/a)1953.55 (n/a)1775.47 (n/a)81.57 (n/a)

gemm_192x384x64_48x96x16_4cols

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:252.05 (n/a)1.66 (n/a)1.86 (n/a)0.79 (n/a)0.53 (n/a)281.60 (n/a)151.72 (n/a)118.60 (n/a)107.90 (n/a)73.91 (n/a)87.43 (n/a)70.78 (n/a)79.55 (n/a)33.52 (n/a)22.49 (n/a)

gemm_192x384x64_48x96x16_4cols_bcolmaj_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:252.23 (n/a)1.82 (n/a)1.87 (n/a)1.32 (n/a)0.33 (n/a)167.80 (n/a)124.82 (n/a)118.30 (n/a)99.10 (n/a)25.58 (n/a)95.25 (n/a)77.86 (n/a)79.77 (n/a)56.23 (n/a)13.96 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.72 (-0.11%)3.55 (-2.75%)3.56 (-2.75%)3.36 (-5.33%)0.12 (+76.10%)2725.23 (-0.11%)2602.52 (-2.75%)2612.11 (-2.75%)2459.89 (-5.33%)86.27 (+76.10%)6984.00 (+5.63%)6608.60 (+2.91%)6577.00 (+2.83%)6304.00 (+0.11%)221.87 (+87.15%)
9e98049 — 2025-11-25 20:50:553.72 (n/a)3.65 (n/a)3.66 (n/a)3.54 (n/a)0.07 (n/a)2728.26 (n/a)2676.06 (n/a)2686.03 (n/a)2598.29 (n/a)48.99 (n/a)6612.00 (n/a)6422.00 (n/a)6396.00 (n/a)6297.00 (n/a)118.55 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_0_bcolmaj_1_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.40 (-3.14%)3.38 (-0.82%)3.38 (-0.82%)3.35 (+4.93%)0.02 (-84.51%)2493.81 (-3.14%)2475.27 (-0.82%)2479.06 (-0.82%)2453.92 (+4.93%)13.33 (-84.51%)7001.00 (-4.70%)6940.80 (+0.71%)6930.00 (+0.83%)6889.00 (+3.24%)37.43 (-84.76%)
9e98049 — 2025-11-25 20:50:553.51 (n/a)3.40 (n/a)3.41 (n/a)3.19 (n/a)0.12 (n/a)2574.53 (n/a)2495.72 (n/a)2499.62 (n/a)2338.67 (n/a)86.02 (n/a)7346.00 (n/a)6892.20 (n/a)6873.00 (n/a)6673.00 (n/a)245.57 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_1_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.02 (+0.52%)2.87 (-2.05%)2.86 (-2.24%)2.70 (-6.16%)0.11 (+131.66%)2216.19 (+0.52%)2104.44 (-2.05%)2094.08 (-2.24%)1980.84 (-6.16%)80.25 (+131.67%)8673.00 (+6.56%)8175.60 (+2.21%)8204.00 (+2.29%)7752.00 (-0.51%)314.03 (+145.36%)
9e98049 — 2025-11-25 20:50:553.01 (n/a)2.93 (n/a)2.92 (n/a)2.88 (n/a)0.05 (n/a)2204.81 (n/a)2148.41 (n/a)2142.13 (n/a)2110.81 (n/a)34.64 (n/a)8139.00 (n/a)7998.60 (n/a)8020.00 (n/a)7792.00 (n/a)127.99 (n/a)

gemm_2048x2048x2048_64x64x64_1cols

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.52 (n/a)0.52 (n/a)0.52 (n/a)0.52 (n/a)0.00 (n/a)48401.80 (n/a)48318.62 (n/a)48391.10 (n/a)48013.20 (n/a)170.80 (n/a)357.82 (n/a)355.56 (n/a)355.02 (n/a)354.94 (n/a)1.26 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:491.04 (+0.94%)1.03 (+0.23%)1.03 (+0.09%)1.02 (+0.07%)0.00 (+197.51%)24576.00 (-0.07%)24485.82 (-0.23%)24530.40 (-0.09%)24266.40 (-0.94%)111.26 (+194.35%)
0a6c11c — 2025-12-04 00:39:101.03 (n/a)1.03 (n/a)1.03 (n/a)1.02 (n/a)0.00 (n/a)24594.30 (n/a)24542.14 (n/a)24551.90 (n/a)24495.50 (n/a)37.80 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:051.04 (+0.28%)1.03 (-0.15%)1.03 (+0.05%)1.02 (-0.52%)0.01 (+42.72%)24719.80 (+0.52%)24512.42 (+0.16%)24529.20 (-0.05%)24092.60 (-0.28%)252.92 (+43.15%)713.08 (+0.28%)700.92 (-0.15%)700.38 (+0.05%)694.98 (-0.52%)7.30 (+42.71%)
50ad37c — 2025-12-29 20:35:531.04 (n/a)1.03 (n/a)1.03 (n/a)1.02 (n/a)0.01 (n/a)24592.30 (n/a)24473.70 (n/a)24540.90 (n/a)24160.80 (n/a)176.68 (n/a)711.06 (n/a)702.00 (n/a)700.05 (n/a)698.59 (n/a)5.12 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_1_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:490.99 (+1.69%)0.97 (+0.29%)0.97 (-0.04%)0.97 (-0.15%)0.01 (+434.92%)26025.30 (+0.15%)25839.34 (-0.29%)25916.70 (+0.04%)25446.00 (-1.67%)202.05 (+425.44%)
0a6c11c — 2025-12-04 00:39:100.97 (n/a)0.97 (n/a)0.97 (n/a)0.97 (n/a)0.00 (n/a)25986.90 (n/a)25913.52 (n/a)25905.70 (n/a)25877.00 (n/a)38.45 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_1_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:050.97 (-0.01%)0.97 (+0.03%)0.97 (+0.04%)0.97 (+0.00%)0.00 (-6.43%)25927.90 (-0.00%)25903.20 (-0.03%)25898.60 (-0.04%)25892.80 (+0.01%)14.05 (-6.56%)663.50 (-0.01%)663.23 (+0.03%)663.35 (+0.04%)662.60 (+0.00%)0.36 (-6.43%)
50ad37c — 2025-12-29 20:35:530.97 (n/a)0.97 (n/a)0.97 (n/a)0.97 (n/a)0.00 (n/a)25928.10 (n/a)25910.14 (n/a)25908.10 (n/a)25891.20 (n/a)15.04 (n/a)663.54 (n/a)663.06 (n/a)663.11 (n/a)662.60 (n/a)0.38 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_1_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:490.92 (-1.02%)0.92 (-0.66%)0.92 (-0.10%)0.91 (-0.78%)0.01 (-4.54%)27751.40 (+0.79%)27480.74 (+0.66%)27354.80 (+0.10%)27320.50 (+1.03%)182.69 (-2.84%)
0a6c11c — 2025-12-04 00:39:100.93 (n/a)0.92 (n/a)0.92 (n/a)0.91 (n/a)0.01 (n/a)27534.80 (n/a)27299.32 (n/a)27326.70 (n/a)27042.30 (n/a)188.04 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_1_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:050.92 (+0.11%)0.91 (-0.12%)0.91 (-0.21%)0.91 (-0.24%)0.00 (+30.34%)27723.50 (+0.24%)27548.94 (+0.12%)27522.90 (+0.21%)27403.60 (-0.11%)124.26 (+30.58%)626.92 (+0.11%)623.62 (-0.12%)624.20 (-0.21%)619.69 (-0.24%)2.81 (+30.34%)
50ad37c — 2025-12-29 20:35:530.92 (n/a)0.91 (n/a)0.92 (n/a)0.91 (n/a)0.00 (n/a)27657.10 (n/a)27515.58 (n/a)27464.70 (n/a)27434.60 (n/a)95.16 (n/a)626.21 (n/a)624.37 (n/a)625.52 (n/a)621.17 (n/a)2.16 (n/a)

gemm_2048x2048x2048_64x64x64_2cols_bcolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.91 (n/a)0.90 (n/a)0.91 (n/a)0.88 (n/a)0.01 (n/a)28447.10 (n/a)27899.18 (n/a)27774.40 (n/a)27519.50 (n/a)359.64 (n/a)624.28 (n/a)615.87 (n/a)618.55 (n/a)603.92 (n/a)7.88 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
31ab24e — 2025-11-22 02:01:563.65 (-1.06%)3.59 (+0.83%)3.63 (-0.26%)3.45 (+2.66%)0.07 (-44.71%)2678.08 (-1.06%)2632.13 (+0.83%)2658.60 (-0.26%)2528.31 (+2.66%)53.86 (-44.71%)6795.00 (-2.59%)6529.80 (-0.92%)6462.00 (+0.26%)6415.00 (+1.07%)137.14 (-45.27%)
d9a64c5 — 2025-11-21 20:49:353.69 (n/a)3.56 (n/a)3.64 (n/a)3.36 (n/a)0.13 (n/a)2706.77 (n/a)2610.50 (n/a)2665.61 (n/a)2462.71 (n/a)97.41 (n/a)6976.00 (n/a)6590.40 (n/a)6445.00 (n/a)6347.00 (n/a)250.58 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.98 (+0.20%)3.84 (-2.95%)3.76 (-5.27%)3.72 (-4.97%)0.13 (+473.97%)6760.00 (+5.23%)6559.50 (+3.13%)6697.10 (+5.56%)6321.00 (-0.20%)218.68 (+499.40%)2717.91 (+0.20%)2621.44 (-2.95%)2565.27 (-5.27%)2541.40 (-4.97%)88.37 (+473.97%)
331dcca — 2026-01-08 18:34:103.97 (n/a)3.96 (n/a)3.97 (n/a)3.92 (n/a)0.02 (n/a)6423.80 (n/a)6360.32 (n/a)6344.30 (n/a)6333.90 (n/a)36.48 (n/a)2712.39 (n/a)2701.17 (n/a)2707.91 (n/a)2674.40 (n/a)15.40 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_1_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.75 (+0.28%)3.70 (+1.25%)3.71 (-0.21%)3.65 (+4.77%)0.04 (-66.90%)6901.80 (-4.56%)6795.46 (-1.30%)6790.20 (+0.21%)6715.90 (-0.28%)67.05 (-68.40%)2558.08 (+0.28%)2528.33 (+1.25%)2530.10 (-0.21%)2489.18 (+4.77%)24.81 (-66.90%)
331dcca — 2026-01-08 18:34:103.74 (n/a)3.66 (n/a)3.71 (n/a)3.48 (n/a)0.11 (n/a)7231.20 (n/a)6885.18 (n/a)6775.70 (n/a)6734.90 (n/a)212.21 (n/a)2550.87 (n/a)2497.04 (n/a)2535.50 (n/a)2375.78 (n/a)74.96 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_1_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.62 (-0.77%)3.54 (+0.14%)3.60 (-0.04%)3.44 (+2.17%)0.09 (-25.77%)7307.40 (-2.13%)7103.46 (-0.18%)6982.50 (+0.04%)6948.20 (+0.77%)182.95 (-26.55%)2472.55 (-0.77%)2419.80 (+0.14%)2460.41 (-0.04%)2351.04 (+2.17%)61.76 (-25.77%)
331dcca — 2026-01-08 18:34:103.65 (n/a)3.54 (n/a)3.61 (n/a)3.37 (n/a)0.12 (n/a)7466.20 (n/a)7116.40 (n/a)6979.60 (n/a)6895.10 (n/a)249.09 (n/a)2491.61 (n/a)2416.45 (n/a)2461.45 (n/a)2301.00 (n/a)83.20 (n/a)

gemm_2048x2048x2048_64x64x64_8cols_bcolmaj_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:253.51 (n/a)3.49 (n/a)3.48 (n/a)3.48 (n/a)0.01 (n/a)7235.40 (n/a)7215.10 (n/a)7227.20 (n/a)7171.80 (n/a)25.51 (n/a)2395.46 (n/a)2381.12 (n/a)2377.13 (n/a)2374.41 (n/a)8.45 (n/a)

gemm_384x1536x1792_32x48x64_4cols_bcolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:254.34 (n/a)4.17 (n/a)4.32 (n/a)3.53 (n/a)0.36 (n/a)2284.00 (n/a)1947.78 (n/a)1866.50 (n/a)1856.20 (n/a)188.02 (n/a)1138.86 (n/a)1092.52 (n/a)1132.55 (n/a)925.53 (n/a)93.39 (n/a)

gemm_896x1792x640_32x64x80_8cols_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:256.72 (n/a)6.31 (n/a)6.61 (n/a)5.19 (n/a)0.65 (n/a)1280.80 (n/a)1064.58 (n/a)1005.80 (n/a)989.40 (n/a)123.84 (n/a)2077.17 (n/a)1949.09 (n/a)2043.44 (n/a)1604.60 (n/a)199.72 (n/a)

layer_norm_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)149.70 (n/a)105.32 (n/a)94.70 (n/a)78.40 (n/a)31.11 (n/a)

layer_norm_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.11 (n/a)0.06 (n/a)0.02 (n/a)139.70 (n/a)94.94 (n/a)77.80 (n/a)74.90 (n/a)28.23 (n/a)

layer_norm_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.10 (n/a)0.06 (n/a)0.02 (n/a)136.20 (n/a)95.34 (n/a)85.40 (n/a)77.80 (n/a)23.63 (n/a)

layer_norm_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.08 (n/a)0.09 (n/a)0.04 (n/a)0.03 (n/a)215.00 (n/a)113.50 (n/a)94.50 (n/a)76.00 (n/a)57.35 (n/a)

layer_norm_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.09 (n/a)0.04 (n/a)0.02 (n/a)189.50 (n/a)106.80 (n/a)90.70 (n/a)78.00 (n/a)46.53 (n/a)

layer_norm_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)179.40 (n/a)109.66 (n/a)93.80 (n/a)85.40 (n/a)39.25 (n/a)

layer_norm_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.08 (n/a)0.08 (n/a)0.05 (n/a)0.02 (n/a)157.30 (n/a)113.02 (n/a)101.40 (n/a)84.90 (n/a)27.83 (n/a)

layer_norm_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)198.00 (n/a)121.56 (n/a)98.80 (n/a)96.60 (n/a)43.50 (n/a)

matrix_vector_mul_128x128_32_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:490.37 (+17.13%)0.30 (+24.87%)0.30 (+38.03%)0.23 (+12.47%)0.06 (+35.23%)
0a6c11c — 2025-12-04 00:39:100.32 (n/a)0.24 (n/a)0.22 (n/a)0.20 (n/a)0.04 (n/a)

matrix_vector_mul_128x128_32_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:100.26 (-1.81%)0.24 (+4.06%)0.24 (+0.52%)0.21 (+16.97%)0.02 (-48.16%)0.26 (-1.81%)0.24 (+4.06%)0.24 (+0.52%)0.20 (+16.97%)0.02 (-48.16%)
a4b6ffe — 2026-01-05 19:24:050.26 (n/a)0.23 (n/a)0.24 (n/a)0.18 (n/a)0.04 (n/a)0.26 (n/a)0.23 (n/a)0.24 (n/a)0.17 (n/a)0.04 (n/a)

matrix_vector_mul_128x128_32tsi_128tso_1col

No metrics available.

matrix_vector_mul_128x128_32tsi_128tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:250.37 (+5.44%)0.31 (-0.05%)0.34 (-1.99%)0.22 (-7.87%)0.06 (+15.13%)0.36 (+5.44%)0.31 (-0.05%)0.34 (-1.99%)0.21 (-7.87%)0.06 (+15.13%)
84d3478 — 2026-02-17 23:25:130.35 (n/a)0.31 (n/a)0.35 (n/a)0.23 (n/a)0.05 (n/a)0.35 (n/a)0.31 (n/a)0.34 (n/a)0.23 (n/a)0.05 (n/a)

matrix_vector_mul_2048x8192_1_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4913.57 (+3.47%)13.15 (+1.09%)13.10 (+0.16%)12.52 (-1.58%)0.37 (+153.19%)
0a6c11c — 2025-12-04 00:39:1013.11 (n/a)13.01 (n/a)13.07 (n/a)12.72 (n/a)0.15 (n/a)

matrix_vector_mul_2048x8192_1_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1013.43 (-0.07%)13.30 (+4.62%)13.27 (+1.29%)13.16 (+16.70%)0.12 (-86.69%)13.42 (-0.07%)13.29 (+4.62%)13.26 (+1.29%)13.16 (+16.70%)0.12 (-86.69%)
a4b6ffe — 2026-01-05 19:24:0513.44 (n/a)12.71 (n/a)13.10 (n/a)11.28 (n/a)0.87 (n/a)13.43 (n/a)12.70 (n/a)13.09 (n/a)11.27 (n/a)0.87 (n/a)

matrix_vector_mul_2048x8192_1_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4925.06 (+2.47%)23.71 (+1.81%)24.13 (+0.50%)20.90 (+5.41%)1.48 (-15.60%)
0a6c11c — 2025-12-04 00:39:1024.46 (n/a)23.29 (n/a)24.01 (n/a)19.83 (n/a)1.75 (n/a)

matrix_vector_mul_2048x8192_1_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1024.46 (-0.60%)21.07 (-8.03%)19.75 (-17.83%)18.08 (+1.18%)3.03 (+7.07%)24.44 (-0.60%)21.06 (-8.03%)19.74 (-17.83%)18.07 (+1.18%)3.03 (+7.07%)
a4b6ffe — 2026-01-05 19:24:0524.60 (n/a)22.91 (n/a)24.04 (n/a)17.87 (n/a)2.83 (n/a)24.59 (n/a)22.90 (n/a)24.02 (n/a)17.86 (n/a)2.83 (n/a)

matrix_vector_mul_2048x8192_1_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4942.05 (+1.54%)40.50 (+12.60%)40.49 (+2.36%)38.98 (+97.53%)0.97 (-88.09%)
0a6c11c — 2025-12-04 00:39:1041.41 (n/a)35.97 (n/a)39.55 (n/a)19.73 (n/a)8.15 (n/a)

matrix_vector_mul_2048x8192_1_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1042.58 (+1.51%)41.37 (+14.29%)41.25 (+4.89%)40.33 (+91.42%)0.92 (-89.18%)42.56 (+1.51%)41.34 (+14.29%)41.23 (+4.89%)40.30 (+91.42%)0.92 (-89.18%)
a4b6ffe — 2026-01-05 19:24:0541.95 (n/a)36.20 (n/a)39.33 (n/a)21.07 (n/a)8.54 (n/a)41.92 (n/a)36.17 (n/a)39.31 (n/a)21.05 (n/a)8.53 (n/a)

matrix_vector_mul_2048x8192_1_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4946.62 (+3.11%)44.82 (+0.17%)44.78 (+0.17%)42.60 (-3.80%)1.60 (+367.72%)
0a6c11c — 2025-12-04 00:39:1045.21 (n/a)44.75 (n/a)44.70 (n/a)44.29 (n/a)0.34 (n/a)

matrix_vector_mul_2048x8192_1_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1046.04 (+4.14%)45.06 (+3.47%)44.55 (+2.76%)44.30 (+2.84%)0.87 (+66.58%)46.01 (+4.14%)45.03 (+3.47%)44.52 (+2.76%)44.27 (+2.84%)0.87 (+66.58%)
a4b6ffe — 2026-01-05 19:24:0544.21 (n/a)43.55 (n/a)43.35 (n/a)43.07 (n/a)0.52 (n/a)44.18 (n/a)43.52 (n/a)43.32 (n/a)43.05 (n/a)0.52 (n/a)

matrix_vector_mul_2048x8192_1tsi_1024tso_2col

No metrics available.

matrix_vector_mul_2048x8192_1tsi_1024tso_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2524.47 (-3.32%)24.16 (-3.13%)24.20 (-3.09%)23.84 (-3.04%)0.25 (-14.36%)24.46 (-3.32%)24.15 (-3.13%)24.18 (-3.09%)23.83 (-3.04%)0.25 (-14.36%)
84d3478 — 2026-02-17 23:25:1325.31 (n/a)24.94 (n/a)24.97 (n/a)24.59 (n/a)0.29 (n/a)25.30 (n/a)24.93 (n/a)24.95 (n/a)24.58 (n/a)0.29 (n/a)

matrix_vector_mul_2048x8192_1tsi_2048tso_1col

No metrics available.

matrix_vector_mul_2048x8192_1tsi_2048tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2513.25 (-3.04%)13.02 (-2.92%)13.10 (-2.64%)12.56 (-4.14%)0.27 (+32.55%)13.24 (-3.04%)13.01 (-2.92%)13.09 (-2.64%)12.55 (-4.14%)0.27 (+32.55%)
84d3478 — 2026-02-17 23:25:1313.66 (n/a)13.41 (n/a)13.45 (n/a)13.10 (n/a)0.20 (n/a)13.65 (n/a)13.40 (n/a)13.44 (n/a)13.09 (n/a)0.20 (n/a)

matrix_vector_mul_2048x8192_1tsi_256tso_8col

No metrics available.

matrix_vector_mul_2048x8192_1tsi_256tso_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2547.71 (+3.51%)45.38 (+9.55%)45.22 (-1.27%)43.47 (+75.24%)1.52 (-83.67%)47.68 (+3.51%)45.35 (+9.55%)45.19 (-1.27%)43.45 (+75.24%)1.52 (-83.67%)
84d3478 — 2026-02-17 23:25:1346.09 (n/a)41.43 (n/a)45.80 (n/a)24.81 (n/a)9.31 (n/a)46.06 (n/a)41.40 (n/a)45.77 (n/a)24.79 (n/a)9.30 (n/a)

matrix_vector_mul_2048x8192_1tsi_512tso_4col

No metrics available.

matrix_vector_mul_2048x8192_1tsi_512tso_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2540.84 (-0.32%)39.53 (-1.25%)39.40 (-1.97%)38.87 (-0.69%)0.77 (-6.24%)40.81 (-0.32%)39.50 (-1.25%)39.37 (-1.97%)38.84 (-0.69%)0.77 (-6.24%)
84d3478 — 2026-02-17 23:25:1340.97 (n/a)40.03 (n/a)40.19 (n/a)39.14 (n/a)0.82 (n/a)40.95 (n/a)40.00 (n/a)40.17 (n/a)39.11 (n/a)0.82 (n/a)

matrix_vector_mul_8192x2048_4_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4913.29 (+1.63%)13.19 (+6.08%)13.19 (+1.17%)13.09 (+26.86%)0.08 (-92.94%)
0a6c11c — 2025-12-04 00:39:1013.08 (n/a)12.44 (n/a)13.03 (n/a)10.32 (n/a)1.07 (n/a)

matrix_vector_mul_8192x2048_4_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1013.68 (+3.42%)13.30 (+8.54%)13.38 (+7.25%)12.72 (+14.96%)0.35 (-66.09%)13.67 (+3.42%)13.29 (+8.54%)13.38 (+7.25%)12.71 (+14.96%)0.35 (-66.09%)
a4b6ffe — 2026-01-05 19:24:0513.23 (n/a)12.25 (n/a)12.48 (n/a)11.07 (n/a)1.04 (n/a)13.22 (n/a)12.24 (n/a)12.47 (n/a)11.06 (n/a)1.04 (n/a)

matrix_vector_mul_8192x2048_4_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4925.58 (+4.24%)24.23 (+6.59%)24.03 (+0.29%)23.58 (+37.29%)0.70 (-74.97%)
0a6c11c — 2025-12-04 00:39:1024.54 (n/a)22.73 (n/a)23.96 (n/a)17.18 (n/a)2.79 (n/a)

matrix_vector_mul_8192x2048_4_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1025.08 (+1.32%)24.68 (+13.42%)24.61 (+3.26%)24.47 (+48.32%)0.24 (-93.56%)25.07 (+1.32%)24.67 (+13.42%)24.59 (+3.26%)24.46 (+48.32%)0.24 (-93.56%)
a4b6ffe — 2026-01-05 19:24:0524.76 (n/a)21.76 (n/a)23.83 (n/a)16.50 (n/a)3.74 (n/a)24.74 (n/a)21.75 (n/a)23.82 (n/a)16.49 (n/a)3.73 (n/a)

matrix_vector_mul_8192x2048_4_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4940.94 (-2.19%)39.74 (-1.10%)39.76 (-0.60%)38.17 (-2.14%)1.04 (+4.39%)
0a6c11c — 2025-12-04 00:39:1041.86 (n/a)40.18 (n/a)40.00 (n/a)39.00 (n/a)1.00 (n/a)

matrix_vector_mul_8192x2048_4_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1040.74 (-1.09%)37.16 (-7.15%)40.25 (+1.47%)24.51 (-37.44%)7.07 (+736.13%)40.72 (-1.09%)37.13 (-7.15%)40.23 (+1.47%)24.50 (-37.44%)7.07 (+736.13%)
a4b6ffe — 2026-01-05 19:24:0541.19 (n/a)40.02 (n/a)39.67 (n/a)39.18 (n/a)0.85 (n/a)41.17 (n/a)39.99 (n/a)39.64 (n/a)39.16 (n/a)0.85 (n/a)

matrix_vector_mul_8192x2048_4_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4946.71 (-1.04%)44.91 (-2.34%)44.71 (-2.98%)43.80 (-1.44%)0.96 (+6.93%)
0a6c11c — 2025-12-04 00:39:1047.20 (n/a)45.98 (n/a)46.08 (n/a)44.44 (n/a)0.90 (n/a)

matrix_vector_mul_8192x2048_4_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1046.28 (+5.68%)44.89 (+14.59%)44.74 (+4.26%)44.20 (+62.80%)0.84 (-87.96%)46.25 (+5.68%)44.86 (+14.59%)44.71 (+4.26%)44.17 (+62.80%)0.84 (-87.96%)
a4b6ffe — 2026-01-05 19:24:0543.79 (n/a)39.17 (n/a)42.91 (n/a)27.15 (n/a)6.98 (n/a)43.77 (n/a)39.15 (n/a)42.88 (n/a)27.13 (n/a)6.98 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_1col

No metrics available.

matrix_vector_mul_8192x2048_4tsi_1024tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2513.34 (-0.79%)13.20 (-0.36%)13.18 (-1.55%)13.12 (+1.75%)0.08 (-65.93%)13.33 (-0.79%)13.19 (-0.36%)13.17 (-1.55%)13.11 (+1.75%)0.08 (-65.93%)
84d3478 — 2026-02-17 23:25:1313.45 (n/a)13.25 (n/a)13.39 (n/a)12.90 (n/a)0.25 (n/a)13.44 (n/a)13.24 (n/a)13.38 (n/a)12.89 (n/a)0.25 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_2col

No metrics available.

matrix_vector_mul_8192x2048_4tsi_1024tso_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2524.10 (-5.45%)23.07 (-1.85%)24.01 (-2.88%)19.26 (+7.47%)2.13 (-32.15%)24.08 (-5.45%)23.05 (-1.85%)23.99 (-2.88%)19.25 (+7.47%)2.13 (-32.15%)
84d3478 — 2026-02-17 23:25:1325.48 (n/a)23.50 (n/a)24.72 (n/a)17.92 (n/a)3.14 (n/a)25.47 (n/a)23.48 (n/a)24.71 (n/a)17.91 (n/a)3.14 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_4col

No metrics available.

matrix_vector_mul_8192x2048_4tsi_1024tso_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2540.80 (-0.16%)39.58 (-1.93%)39.75 (-1.71%)38.28 (-2.50%)1.10 (+67.33%)40.78 (-0.16%)39.55 (-1.93%)39.73 (-1.71%)38.26 (-2.50%)1.10 (+67.33%)
84d3478 — 2026-02-17 23:25:1340.87 (n/a)40.36 (n/a)40.45 (n/a)39.26 (n/a)0.66 (n/a)40.84 (n/a)40.33 (n/a)40.42 (n/a)39.24 (n/a)0.65 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_8col

No metrics available.

matrix_vector_mul_8192x2048_4tsi_1024tso_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2549.31 (+6.27%)46.12 (+1.67%)45.30 (-0.27%)44.62 (+1.87%)1.85 (+85.10%)49.28 (+6.27%)46.10 (+1.67%)45.28 (-0.27%)44.59 (+1.87%)1.85 (+85.10%)
84d3478 — 2026-02-17 23:25:1346.40 (n/a)45.37 (n/a)45.43 (n/a)43.80 (n/a)1.00 (n/a)46.37 (n/a)45.34 (n/a)45.40 (n/a)43.78 (n/a)1.00 (n/a)

mem_copy_16_cores_2_chans_2048_tile_128_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)173.30 (n/a)137.94 (n/a)133.60 (n/a)123.40 (n/a)18.40 (n/a)

mem_copy_16_cores_2_chans_2048_tile_128_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-2.12%)0.06 (-16.44%)0.06 (-21.48%)0.03 (-23.84%)0.02 (+10.15%)247.60 (+31.28%)148.38 (+24.08%)128.30 (+27.41%)100.10 (+2.14%)58.70 (+50.73%)
84d3478 — 2026-02-17 23:25:130.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)188.60 (n/a)119.58 (n/a)100.70 (n/a)98.00 (n/a)38.94 (n/a)

mem_copy_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (-16.99%)0.05 (-12.47%)0.05 (-7.32%)0.04 (-11.17%)0.00 (-29.61%)197.00 (+12.57%)171.00 (+13.70%)164.00 (+7.89%)153.00 (+20.47%)17.49 (-4.32%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)175.00 (n/a)150.40 (n/a)152.00 (n/a)127.00 (n/a)18.28 (n/a)

mem_copy_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (+5.56%)0.05 (-1.59%)0.05 (-1.39%)0.03 (-14.07%)0.01 (+25.03%)263.00 (+16.37%)166.00 (+4.53%)144.00 (+1.41%)126.00 (-5.26%)49.38 (+43.17%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)226.00 (n/a)158.80 (n/a)142.00 (n/a)133.00 (n/a)34.49 (n/a)

mem_copy_1_cores_1_chans_2048_tile_2048_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.08 (n/a)0.07 (n/a)0.07 (n/a)0.05 (n/a)0.01 (n/a)150.20 (n/a)121.68 (n/a)125.60 (n/a)97.90 (n/a)18.56 (n/a)

mem_copy_1_cores_1_chans_2048_tile_2048_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (-0.24%)0.08 (-17.53%)0.08 (-27.05%)0.06 (-15.43%)0.02 (+4.62%)148.50 (+18.23%)113.02 (+23.38%)103.60 (+37.04%)74.10 (+0.14%)31.69 (+33.33%)
84d3478 — 2026-02-17 23:25:130.11 (n/a)0.09 (n/a)0.11 (n/a)0.07 (n/a)0.02 (n/a)125.60 (n/a)91.60 (n/a)75.60 (n/a)74.00 (n/a)23.77 (n/a)

mem_copy_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (-15.32%)0.05 (-17.09%)0.05 (-23.81%)0.04 (-15.54%)0.01 (-18.18%)193.00 (+18.40%)160.80 (+20.36%)168.00 (+31.25%)124.00 (+18.10%)22.92 (+11.39%)
0999a9a — 2025-11-13 19:24:080.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)163.00 (n/a)133.60 (n/a)128.00 (n/a)105.00 (n/a)20.58 (n/a)

mem_copy_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (-23.02%)0.05 (-6.98%)0.05 (-2.13%)0.04 (-15.31%)0.01 (-33.29%)196.00 (+18.07%)147.00 (+6.52%)141.00 (+2.17%)126.00 (+29.90%)25.27 (+8.13%)
0999a9a — 2025-11-13 19:24:080.08 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)166.00 (n/a)138.00 (n/a)138.00 (n/a)97.00 (n/a)23.37 (n/a)

mem_copy_2_cores_1_chans_2048_tile_1024_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.10 (n/a)0.08 (n/a)0.08 (n/a)0.07 (n/a)0.01 (n/a)109.90 (n/a)98.30 (n/a)100.80 (n/a)79.40 (n/a)10.54 (n/a)

mem_copy_2_cores_1_chans_2048_tile_1024_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (+27.03%)0.09 (+25.79%)0.11 (+27.09%)0.05 (+12.74%)0.03 (+43.93%)161.90 (-11.29%)94.34 (-18.35%)76.30 (-21.34%)74.00 (-21.36%)38.01 (+0.49%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)182.50 (n/a)115.54 (n/a)97.00 (n/a)94.10 (n/a)37.82 (n/a)

mem_copy_2_cores_2_chans_2048_tile_1024_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.10 (n/a)0.07 (n/a)0.07 (n/a)0.05 (n/a)0.01 (n/a)150.10 (n/a)114.60 (n/a)111.70 (n/a)83.90 (n/a)21.23 (n/a)

mem_copy_2_cores_2_chans_2048_tile_1024_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.10 (+2.70%)0.08 (-5.38%)0.09 (+2.10%)0.04 (-43.30%)0.02 (+170.50%)201.60 (+76.38%)112.08 (+16.53%)90.80 (-2.05%)84.60 (-2.65%)50.14 (+375.75%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.09 (n/a)0.09 (n/a)0.07 (n/a)0.01 (n/a)114.30 (n/a)96.18 (n/a)92.70 (n/a)86.90 (n/a)10.54 (n/a)

mem_copy_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.07 (+11.97%)0.05 (+3.09%)0.05 (+2.10%)0.03 (-11.35%)0.01 (+41.65%)229.00 (+12.81%)156.00 (-0.51%)143.00 (-2.05%)117.00 (-10.69%)38.21 (+48.71%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)203.00 (n/a)156.80 (n/a)146.00 (n/a)131.00 (n/a)25.69 (n/a)

mem_copy_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (+9.02%)0.05 (+5.59%)0.05 (+0.00%)0.04 (+3.63%)0.01 (+25.48%)193.00 (-3.50%)147.40 (-4.66%)147.00 (+0.00%)122.00 (-8.27%)25.24 (+7.48%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)200.00 (n/a)154.60 (n/a)147.00 (n/a)133.00 (n/a)23.48 (n/a)

mem_copy_4_cores_1_chans_2048_tile_512_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.09 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)129.70 (n/a)113.30 (n/a)116.90 (n/a)90.40 (n/a)15.04 (n/a)

mem_copy_4_cores_1_chans_2048_tile_512_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.09 (-5.87%)0.08 (-3.61%)0.08 (-0.79%)0.06 (-12.74%)0.01 (+22.04%)146.30 (+14.66%)108.02 (+4.96%)99.30 (+0.81%)95.10 (+6.26%)21.49 (+48.38%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)127.60 (n/a)102.92 (n/a)98.50 (n/a)89.50 (n/a)14.49 (n/a)

mem_copy_4_cores_2_chans_2048_tile_512_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.08 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)134.50 (n/a)116.64 (n/a)116.80 (n/a)98.00 (n/a)15.71 (n/a)

mem_copy_4_cores_2_chans_2048_tile_512_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (+13.04%)0.09 (+17.19%)0.10 (+20.43%)0.06 (+17.23%)0.02 (+16.34%)142.80 (-14.70%)91.94 (-14.71%)79.00 (-17.02%)77.70 (-11.50%)28.50 (-14.82%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)167.40 (n/a)107.80 (n/a)95.20 (n/a)87.80 (n/a)33.46 (n/a)

mem_copy_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (-13.48%)0.05 (-17.79%)0.05 (-15.33%)0.03 (-25.09%)0.01 (+6.67%)267.00 (+33.50%)176.80 (+24.51%)150.00 (+18.11%)141.00 (+15.57%)47.41 (+61.41%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.06 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)200.00 (n/a)142.00 (n/a)127.00 (n/a)122.00 (n/a)29.37 (n/a)

mem_copy_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (+13.38%)0.05 (+23.85%)0.05 (+37.91%)0.05 (+30.72%)0.00 (-56.85%)166.00 (-23.50%)153.40 (-20.60%)153.00 (-27.49%)142.00 (-11.80%)7.63 (-70.60%)
0999a9a — 2025-11-13 19:24:080.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)217.00 (n/a)193.20 (n/a)211.00 (n/a)161.00 (n/a)25.96 (n/a)

mem_copy_8_cores_1_chans_2048_tile_256_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.09 (n/a)0.08 (n/a)0.08 (n/a)0.07 (n/a)0.01 (n/a)117.30 (n/a)107.80 (n/a)108.50 (n/a)94.10 (n/a)8.29 (n/a)

mem_copy_8_cores_1_chans_2048_tile_256_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-0.39%)0.07 (+10.07%)0.08 (+1.07%)0.06 (+38.38%)0.01 (-47.35%)136.70 (-27.75%)110.64 (-13.07%)104.70 (-1.13%)103.00 (+0.39%)14.59 (-60.78%)
84d3478 — 2026-02-17 23:25:130.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)189.20 (n/a)127.28 (n/a)105.90 (n/a)102.60 (n/a)37.20 (n/a)

mem_copy_8_cores_2_chans_2048_tile_256_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.09 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)136.50 (n/a)120.86 (n/a)125.10 (n/a)91.60 (n/a)15.72 (n/a)

mem_copy_8_cores_2_chans_2048_tile_256_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-17.85%)0.07 (-24.66%)0.08 (-18.40%)0.04 (-32.08%)0.02 (+21.78%)185.30 (+47.18%)127.72 (+37.87%)102.00 (+22.45%)100.40 (+21.70%)38.66 (+106.34%)
84d3478 — 2026-02-17 23:25:130.10 (n/a)0.09 (n/a)0.10 (n/a)0.07 (n/a)0.01 (n/a)125.90 (n/a)92.64 (n/a)83.30 (n/a)82.50 (n/a)18.74 (n/a)

mha

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47253.90 (n/a)47213.70 (n/a)47210.90 (n/a)47167.70 (n/a)28.71 (n/a)

mha0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
84d3478 — 2026-02-17 23:25:130.18 (-0.51%)0.18 (-0.36%)0.18 (-0.41%)0.18 (-0.21%)0.00 (-35.21%)47468.60 (+0.21%)47403.94 (+0.36%)47426.00 (+0.41%)47314.40 (+0.51%)68.83 (-34.73%)
1531781 — 2026-01-23 18:49:240.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47367.50 (n/a)47232.68 (n/a)47232.00 (n/a)47073.00 (n/a)105.45 (n/a)

mha_16384_64_1_8_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47533.70 (n/a)47380.56 (n/a)47337.90 (n/a)47268.00 (n/a)110.60 (n/a)

relu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.07 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)151.00 (n/a)115.64 (n/a)93.40 (n/a)92.00 (n/a)31.56 (n/a)

relu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.09 (n/a)0.06 (n/a)0.01 (n/a)130.70 (n/a)103.64 (n/a)94.20 (n/a)93.30 (n/a)16.06 (n/a)

relu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.10 (n/a)0.05 (n/a)0.03 (n/a)172.70 (n/a)97.40 (n/a)78.30 (n/a)75.60 (n/a)42.18 (n/a)

relu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)202.50 (n/a)123.04 (n/a)103.30 (n/a)101.30 (n/a)44.44 (n/a)

rms_norm_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)142.90 (n/a)109.24 (n/a)97.30 (n/a)96.20 (n/a)20.13 (n/a)

rms_norm_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)146.80 (n/a)108.12 (n/a)98.20 (n/a)96.90 (n/a)21.70 (n/a)

rms_norm_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.12 (n/a)0.10 (n/a)0.10 (n/a)0.07 (n/a)0.02 (n/a)125.20 (n/a)85.34 (n/a)78.00 (n/a)68.10 (n/a)22.68 (n/a)

rms_norm_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)140.00 (n/a)106.00 (n/a)97.50 (n/a)96.70 (n/a)19.02 (n/a)

rms_norm_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.01 (n/a)152.20 (n/a)106.26 (n/a)95.60 (n/a)93.30 (n/a)25.71 (n/a)

rms_norm_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)130.40 (n/a)103.48 (n/a)97.50 (n/a)95.70 (n/a)15.08 (n/a)

rms_norm_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)190.30 (n/a)119.44 (n/a)103.80 (n/a)95.20 (n/a)39.79 (n/a)

rms_norm_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.10 (n/a)0.04 (n/a)0.03 (n/a)196.90 (n/a)104.90 (n/a)83.10 (n/a)77.80 (n/a)51.49 (n/a)

rope_1_cols_2_channels_4096_tile_4096_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.19 (-21.66%)0.17 (-8.53%)0.18 (+2.09%)0.14 (-6.71%)0.02 (-38.07%)174.10 (+7.20%)150.06 (+8.11%)139.60 (-2.04%)129.80 (+27.63%)20.36 (-10.65%)
d48746f — 2026-01-12 22:17:100.24 (n/a)0.18 (n/a)0.17 (n/a)0.15 (n/a)0.04 (n/a)162.40 (n/a)138.80 (n/a)142.50 (n/a)101.70 (n/a)22.79 (n/a)

rope_1c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:251.06 (n/a)0.89 (n/a)0.91 (n/a)0.71 (n/a)0.15 (n/a)138.30 (n/a)113.54 (n/a)108.10 (n/a)92.60 (n/a)19.18 (n/a)

rope_1c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.62 (n/a)0.50 (n/a)0.48 (n/a)0.44 (n/a)0.07 (n/a)165.70 (n/a)148.42 (n/a)152.30 (n/a)119.70 (n/a)17.10 (n/a)

rope_2_cols_2_channels_4096_tile_2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.21 (-0.56%)0.17 (-10.62%)0.16 (-13.58%)0.13 (-13.33%)0.03 (+17.05%)183.70 (+15.39%)151.76 (+12.77%)155.20 (+15.73%)117.30 (+0.60%)24.24 (+35.73%)
d48746f — 2026-01-12 22:17:100.21 (n/a)0.19 (n/a)0.18 (n/a)0.15 (n/a)0.02 (n/a)159.20 (n/a)134.58 (n/a)134.10 (n/a)116.60 (n/a)17.86 (n/a)

rope_2c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:251.08 (n/a)0.90 (n/a)0.93 (n/a)0.57 (n/a)0.20 (n/a)172.50 (n/a)115.64 (n/a)105.90 (n/a)91.00 (n/a)33.09 (n/a)

rope_2c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.88 (n/a)0.63 (n/a)0.51 (n/a)0.49 (n/a)0.18 (n/a)151.50 (n/a)124.06 (n/a)143.70 (n/a)84.20 (n/a)30.96 (n/a)

rope_4_cols_2_channels_4096_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.22 (+18.92%)0.17 (+6.24%)0.17 (+5.57%)0.14 (-0.50%)0.03 (+90.77%)171.10 (+0.47%)147.40 (-4.39%)148.10 (-5.31%)110.10 (-15.89%)23.59 (+60.67%)
d48746f — 2026-01-12 22:17:100.19 (n/a)0.16 (n/a)0.16 (n/a)0.14 (n/a)0.02 (n/a)170.30 (n/a)154.16 (n/a)156.40 (n/a)130.90 (n/a)14.69 (n/a)

rope_8_cols_2_channels_4096_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.17 (-1.30%)0.15 (+6.63%)0.15 (-8.78%)0.14 (+39.68%)0.01 (-67.35%)174.90 (-28.41%)163.42 (-10.36%)167.40 (+9.63%)146.50 (+1.31%)11.19 (-76.17%)
d48746f — 2026-01-12 22:17:100.17 (n/a)0.14 (n/a)0.16 (n/a)0.10 (n/a)0.03 (n/a)244.30 (n/a)182.30 (n/a)152.70 (n/a)144.60 (n/a)46.98 (n/a)

rope_8c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:251.07 (n/a)0.88 (n/a)1.03 (n/a)0.54 (n/a)0.23 (n/a)182.80 (n/a)119.18 (n/a)95.90 (n/a)92.30 (n/a)38.79 (n/a)

rope_8c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.56 (n/a)0.48 (n/a)0.52 (n/a)0.36 (n/a)0.09 (n/a)203.00 (n/a)158.30 (n/a)140.40 (n/a)131.30 (n/a)31.38 (n/a)

sigmoid_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)151.30 (n/a)114.02 (n/a)94.10 (n/a)93.50 (n/a)28.04 (n/a)

sigmoid_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.11 (n/a)0.05 (n/a)0.03 (n/a)176.50 (n/a)106.80 (n/a)76.90 (n/a)76.40 (n/a)44.81 (n/a)

sigmoid_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.08 (n/a)0.05 (n/a)0.02 (n/a)171.60 (n/a)112.92 (n/a)100.20 (n/a)90.30 (n/a)33.29 (n/a)

sigmoid_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.09 (n/a)0.10 (n/a)0.04 (n/a)0.03 (n/a)213.10 (n/a)109.18 (n/a)83.40 (n/a)82.60 (n/a)58.09 (n/a)

silu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.08 (n/a)0.10 (n/a)0.04 (n/a)0.03 (n/a)188.60 (n/a)112.80 (n/a)78.60 (n/a)75.70 (n/a)51.04 (n/a)

silu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.02 (n/a)144.40 (n/a)111.98 (n/a)108.80 (n/a)87.10 (n/a)24.06 (n/a)

silu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.08 (n/a)0.05 (n/a)0.02 (n/a)165.90 (n/a)110.66 (n/a)99.00 (n/a)89.80 (n/a)31.15 (n/a)

silu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.09 (n/a)0.09 (n/a)0.04 (n/a)0.02 (n/a)182.80 (n/a)104.92 (n/a)87.20 (n/a)81.50 (n/a)43.64 (n/a)

softmax_1_cols_2_channels_4096_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.19 (-8.94%)0.17 (-8.66%)0.17 (-19.83%)0.14 (+4.32%)0.02 (-35.13%)115.80 (-4.14%)99.26 (+7.47%)98.60 (+24.65%)85.00 (+9.82%)13.65 (-30.86%)
84d3478 — 2026-02-17 23:25:130.21 (n/a)0.18 (n/a)0.21 (n/a)0.14 (n/a)0.04 (n/a)120.80 (n/a)92.36 (n/a)79.10 (n/a)77.40 (n/a)19.75 (n/a)

softmax_2_cols_2_channels_32768_tile_1024

No metrics available.

softmax_2_cols_2_channels_32768_tile_512

No metrics available.

softmax_2_cols_2_channels_4096_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.20 (-0.87%)0.16 (+6.57%)0.17 (-9.55%)0.11 (+29.81%)0.03 (-44.34%)145.70 (-22.99%)104.68 (-15.00%)95.00 (+10.59%)83.80 (+0.84%)24.16 (-54.23%)
84d3478 — 2026-02-17 23:25:130.20 (n/a)0.15 (n/a)0.19 (n/a)0.09 (n/a)0.06 (n/a)189.20 (n/a)123.16 (n/a)85.90 (n/a)83.10 (n/a)52.80 (n/a)

softmax_2_cols_2_channels_4096_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.25 (+23.53%)0.18 (+9.42%)0.18 (-3.86%)0.11 (+2.99%)0.05 (+19.24%)144.10 (-2.90%)98.98 (-8.18%)92.10 (+4.07%)64.30 (-19.02%)29.58 (-7.13%)
84d3478 — 2026-02-17 23:25:130.21 (n/a)0.16 (n/a)0.19 (n/a)0.11 (n/a)0.04 (n/a)148.40 (n/a)107.80 (n/a)88.50 (n/a)79.40 (n/a)31.85 (n/a)

softmax_4_cols_4_channels_32768_tile_2048

No metrics available.

swiglu

No metrics available.

swiglu_decode_1x2048x2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)5187.20 (n/a)4720.84 (n/a)4750.90 (n/a)4015.10 (n/a)390.23 (n/a)

swiglu_decode_1x2048x2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.00 (-0.21%)0.00 (+7.74%)0.00 (+0.37%)0.00 (+8.84%)0.00 (-19.58%)5170.80 (-8.12%)3761.80 (-8.70%)3408.60 (-0.36%)3407.30 (+0.21%)787.66 (-22.98%)
84d3478 — 2026-02-17 23:25:130.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)5627.70 (n/a)4120.44 (n/a)3421.00 (n/a)3400.30 (n/a)1022.61 (n/a)

swiglu_prefill_256x2048x2048

No metrics available.

tanh_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)129.70 (n/a)106.52 (n/a)105.00 (n/a)95.30 (n/a)13.95 (n/a)

tanh_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.11 (n/a)0.05 (n/a)0.02 (n/a)159.90 (n/a)94.92 (n/a)76.10 (n/a)74.70 (n/a)36.73 (n/a)

tanh_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.07 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)129.00 (n/a)112.60 (n/a)106.60 (n/a)101.30 (n/a)12.97 (n/a)

tanh_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)133.70 (n/a)110.00 (n/a)101.90 (n/a)99.30 (n/a)14.49 (n/a)

transpose_2048_M_64_N_1_cols_1_channels_64_m_64_n_8_s

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:493.93 (n/a)3.68 (n/a)3.66 (n/a)3.45 (n/a)0.15 (n/a)151.80 (n/a)142.70 (n/a)143.40 (n/a)133.50 (n/a)5.90 (n/a)

transpose_2048_M_64_N_1_cols_1_channels_64_m_64_n_8_s0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:254.43 (+0.85%)3.73 (+3.83%)4.02 (-1.30%)2.97 (+16.77%)0.65 (-24.99%)176.70 (-14.35%)144.46 (-6.24%)130.40 (+1.32%)118.40 (-0.84%)26.60 (-35.16%)
84d3478 — 2026-02-17 23:25:134.39 (n/a)3.59 (n/a)4.07 (n/a)2.54 (n/a)0.87 (n/a)206.30 (n/a)154.08 (n/a)128.70 (n/a)119.40 (n/a)41.02 (n/a)

transpose_2048_M_64_N_1_cols_2_channels_64_m_64_n_8_s

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:495.11 (n/a)3.85 (n/a)3.40 (n/a)3.29 (n/a)0.71 (n/a)159.60 (n/a)140.18 (n/a)154.20 (n/a)102.60 (n/a)22.70 (n/a)

transpose_2048_M_64_N_1_cols_2_channels_64_m_64_n_8_s0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:255.27 (-14.18%)4.35 (-2.49%)4.45 (+2.15%)2.78 (-22.18%)0.98 (-3.38%)188.60 (+28.47%)126.98 (+4.20%)117.90 (-2.08%)99.40 (+16.53%)36.02 (+50.15%)
84d3478 — 2026-02-17 23:25:136.14 (n/a)4.46 (n/a)4.35 (n/a)3.57 (n/a)1.02 (n/a)146.80 (n/a)121.86 (n/a)120.40 (n/a)85.30 (n/a)23.99 (n/a)

weighted_rms_norm_1_cols_2_channels_2048_weights_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.07 (n/a)0.07 (n/a)0.05 (n/a)0.01 (n/a)169.30 (n/a)121.18 (n/a)112.90 (n/a)99.90 (n/a)28.51 (n/a)

weighted_rms_norm_2_cols_2_channels_2048_weights_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.10 (n/a)0.10 (n/a)0.06 (n/a)0.02 (n/a)140.20 (n/a)90.80 (n/a)78.80 (n/a)77.90 (n/a)27.62 (n/a)

weighted_rms_norm_4_cols_2_channels_2048_weights_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.10 (n/a)0.10 (n/a)0.06 (n/a)0.02 (n/a)126.70 (n/a)89.18 (n/a)79.70 (n/a)77.20 (n/a)21.10 (n/a)

weighted_rms_norm_8_cols_2_channels_2048_weights_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
197d469 — 2026-03-18 17:00:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.07 (n/a)0.08 (n/a)0.05 (n/a)0.01 (n/a)177.80 (n/a)119.42 (n/a)104.00 (n/a)102.90 (n/a)32.72 (n/a)

@github-actions
Copy link
Contributor

📊 Test Results for Test Example Applications

197d469 (2026_03_18_17_06_05)

IRONCLAD

Tested on 2026_03_18_17_06_05 at commit 197d469.

Test Checks TTFT (mean)TPS (mean)
llama_3.2_1b_prompt_1024_tokens_40 ❌ 0/1 n/a n/a
📈 Trends (vs main branch) for Test Example Applications

197d469 (2026_03_18_17_06_05)

IRONCLAD Trends

llama_3.2_1b

Commit/Date Num Tokens (max)Num Tokens (mean)Num Tokens (median)Num Tokens (min)Num Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)Total (max)Total (mean)Total (median)Total (min)Total (stddev)
130b6ea — 2025-12-05 21:33:1240.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.71 (-0.42%)4.64 (-0.09%)4.64 (+0.65%)4.55 (-0.22%)0.05 (-17.66%)4.41 (-0.34%)4.39 (-0.19%)4.38 (-0.33%)4.37 (-0.15%)0.01 (-25.90%)12.96 (-0.00%)12.80 (+0.07%)12.80 (-0.23%)12.67 (+0.44%)0.09 (-21.12%)
0a6c11c — 2025-12-03 23:35:1540.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.73 (n/a)4.64 (n/a)4.61 (n/a)4.56 (n/a)0.06 (n/a)4.42 (n/a)4.40 (n/a)4.40 (n/a)4.37 (n/a)0.02 (n/a)12.96 (n/a)12.79 (n/a)12.83 (n/a)12.62 (n/a)0.12 (n/a)

llama_3.2_1b_prompt_1024_tokens_40

No metrics available.

llama_3.2_1b_prompt_13_tokens_1

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:071.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.62 (-4.45%)0.62 (-4.45%)0.62 (-4.45%)0.62 (-4.45%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:231.00 (n/a)1.00 (n/a)1.00 (n/a)1.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.65 (n/a)0.65 (n/a)0.65 (n/a)0.65 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_13_tokens_40

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:0740.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.30 (+2.38%)4.30 (+2.38%)4.30 (+2.38%)4.30 (+2.38%)0.00 (n/a)0.61 (-2.49%)0.61 (-2.49%)0.61 (-2.49%)0.61 (-2.49%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:2340.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.20 (n/a)4.20 (n/a)4.20 (n/a)4.20 (n/a)0.00 (n/a)0.63 (n/a)0.63 (n/a)0.63 (n/a)0.63 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_2048_tokens_1

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:071.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:231.00 (n/a)1.00 (n/a)1.00 (n/a)1.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_2048_tokens_40

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:0740.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)0.00 (n/a)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:2340.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)0.00 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)0.00 (n/a)

@github-actions
Copy link
Contributor

📊 Test Results for Small Benchmark/Test Suite

2f954d5 (2026_03_18_18_09_41)

IRONCLAD

Tested on 2026_03_18_18_09_41 at commit 2f954d5.

Test Checks Latency (mean)Bandwidth (mean)GFLOP/s (mean)
axpy_1_cols_2_channels_2048_tile_2048_3.0 ✅ 5/5 154.58 0.08 n/a
axpy_2_cols_2_channels_2048_tile_1024_3.0 ✅ 5/5 178.40 0.07 n/a
axpy_4_cols_2_channels_2048_tile_512_3.0 ✅ 5/5 191.58 0.06 n/a
axpy_8_cols_2_channels_2048_tile_256_3.0 ✅ 5/5 201.28 0.06 n/a
dequant_1_cols_1_channels_2048_tile_2048 ✅ 5/5 180.82 0.03 n/a
dequant_1_cols_2_channels_2048_tile_1024 ✅ 5/5 183.80 0.03 n/a
dequant_2_cols_1_channels_2048_tile_1024 ✅ 5/5 175.54 0.03 n/a
dequant_2_cols_2_channels_2048_tile_512 ✅ 5/5 168.70 0.03 n/a
dequant_4_cols_1_channels_2048_tile_512 ✅ 5/5 176.78 0.03 n/a
dequant_4_cols_2_channels_2048_tile_256 ✅ 5/5 197.04 0.03 n/a
dequant_8_cols_1_channels_2048_tile_256 ✅ 5/5 195.96 0.03 n/a
dequant_8_cols_2_channels_2048_tile_128 ✅ 5/5 220.92 0.02 n/a
eltwise_add_1_cols_2_channels_2048_tile_2048 ✅ 5/5 163.22 0.08 n/a
eltwise_add_2_cols_2_channels_2048_tile_1024 ✅ 5/5 216.50 0.07 n/a
eltwise_add_4_cols_2_channels_2048_tile_512 ✅ 5/5 182.82 0.07 n/a
eltwise_add_8_cols_2_channels_2048_tile_256 ✅ 5/5 203.60 0.06 n/a
eltwise_mul_1_cols_2_channels_2048_tile_2048 ✅ 5/5 153.74 0.09 n/a
eltwise_mul_2_cols_2_channels_2048_tile_1024 ✅ 5/5 188.66 0.07 n/a
eltwise_mul_4_cols_2_channels_2048_tile_512 ✅ 5/5 195.90 0.07 n/a
eltwise_mul_8_cols_2_channels_2048_tile_256 ✅ 5/5 221.76 0.06 n/a
gelu_1_cols_1_channels_2048_tile_2048 ✅ 5/5 186.40 0.05 n/a
gelu_1_cols_2_channels_2048_tile_1024 ✅ 5/5 191.02 0.05 n/a
gelu_2_cols_1_channels_2048_tile_1024 ✅ 5/5 204.08 0.04 n/a
gelu_2_cols_2_channels_2048_tile_512 ✅ 5/5 198.10 0.04 n/a
gelu_4_cols_1_channels_2048_tile_512 ✅ 5/5 203.42 0.04 n/a
gelu_4_cols_2_channels_2048_tile_256 ✅ 5/5 187.24 0.05 n/a
gelu_8_cols_1_channels_2048_tile_256 ✅ 5/5 206.00 0.04 n/a
gelu_8_cols_2_channels_2048_tile_128 ✅ 5/5 233.98 0.04 n/a
gemm_1792x896x1152_64x32x48_8cols_ccolmaj ✅ 5/5 2324.70 4.13 n/a
gemm_192x384x64_48x96x16_4cols ✅ 5/5 367.02 0.65 n/a
gemm_192x384x64_48x96x16_4cols_bcolmaj_ccolmaj ✅ 5/5 269.60 0.89 n/a
gemm_2048x2048x2048_64x64x64_1cols ✅ 5/5 47900.00 0.53 n/a
gemm_2048x2048x2048_64x64x64_2cols_bcolmaj ✅ 5/5 27230.48 0.92 n/a
gemm_2048x2048x2048_64x64x64_8cols_bcolmaj_ccolmaj ✅ 5/5 7785.86 3.23 n/a
gemm_384x1536x1792_32x48x64_4cols_bcolmaj ✅ 5/5 2261.72 3.62 n/a
gemm_896x1792x640_32x64x80_8cols_ccolmaj ✅ 5/5 1756.22 3.90 n/a
layer_norm_1_cols_1_channels_2048_tile_2048 ✅ 5/5 194.70 0.04 n/a
layer_norm_1_cols_2_channels_2048_tile_1024 ✅ 5/5 207.32 0.04 n/a
layer_norm_2_cols_1_channels_2048_tile_1024 ✅ 5/5 184.82 0.05 n/a
layer_norm_2_cols_2_channels_2048_tile_512 ✅ 5/5 231.78 0.04 n/a
layer_norm_4_cols_1_channels_2048_tile_512 ✅ 5/5 212.02 0.04 n/a
layer_norm_4_cols_2_channels_2048_tile_256 ✅ 5/5 199.58 0.04 n/a
layer_norm_8_cols_1_channels_2048_tile_256 ✅ 5/5 249.74 0.03 n/a
layer_norm_8_cols_2_channels_2048_tile_128 ✅ 5/5 275.98 0.03 n/a
matrix_vector_mul_128x128_32tsi_128tso_1col ✅ 5/5 n/a 0.19 n/a
matrix_vector_mul_2048x8192_1tsi_1024tso_2col ✅ 5/5 n/a 23.45 n/a
matrix_vector_mul_2048x8192_1tsi_2048tso_1col ✅ 5/5 n/a 12.96 n/a
matrix_vector_mul_2048x8192_1tsi_256tso_8col ✅ 5/5 n/a 41.35 n/a
matrix_vector_mul_2048x8192_1tsi_512tso_4col ✅ 5/5 n/a 38.87 n/a
matrix_vector_mul_8192x2048_4tsi_1024tso_1col ✅ 5/5 n/a 12.61 n/a
matrix_vector_mul_8192x2048_4tsi_1024tso_2col ✅ 5/5 n/a 23.55 n/a
matrix_vector_mul_8192x2048_4tsi_1024tso_4col ✅ 5/5 n/a 38.47 n/a
matrix_vector_mul_8192x2048_4tsi_1024tso_8col ✅ 5/5 n/a 40.86 n/a
mem_copy_16_cores_2_chans_2048_tile_128_False ✅ 5/5 239.32 0.03 n/a
mem_copy_1_cores_1_chans_2048_tile_2048_False ✅ 5/5 195.22 0.04 n/a
mem_copy_2_cores_1_chans_2048_tile_1024_False ✅ 5/5 187.68 0.05 n/a
mem_copy_2_cores_2_chans_2048_tile_1024_False ✅ 5/5 214.68 0.04 n/a
mem_copy_4_cores_1_chans_2048_tile_512_False ✅ 5/5 200.08 0.04 n/a
mem_copy_4_cores_2_chans_2048_tile_512_False ✅ 5/5 201.32 0.04 n/a
mem_copy_8_cores_1_chans_2048_tile_256_False ✅ 5/5 203.44 0.04 n/a
mem_copy_8_cores_2_chans_2048_tile_256_False ✅ 5/5 221.76 0.04 n/a
mha ✅ 5/5 47490.22 0.18 n/a
relu_1_cols_1_channels_2048_tile_2048 ✅ 5/5 183.78 0.05 n/a
relu_2_cols_1_channels_2048_tile_1024 ✅ 5/5 185.02 0.05 n/a
relu_4_cols_1_channels_2048_tile_512 ✅ 5/5 202.80 0.04 n/a
relu_8_cols_1_channels_2048_tile_256 ✅ 5/5 184.50 0.05 n/a
rms_norm_1_cols_1_channels_2048_tile_2048 ✅ 5/5 180.46 0.05 n/a
rms_norm_1_cols_2_channels_2048_tile_1024 ✅ 5/5 180.46 0.05 n/a
rms_norm_2_cols_1_channels_2048_tile_1024 ✅ 5/5 154.88 0.06 n/a
rms_norm_2_cols_2_channels_2048_tile_512 ✅ 5/5 162.22 0.05 n/a
rms_norm_4_cols_1_channels_2048_tile_512 ✅ 5/5 165.72 0.05 n/a
rms_norm_4_cols_2_channels_2048_tile_256 ✅ 5/5 184.16 0.05 n/a
rms_norm_8_cols_1_channels_2048_tile_256 ✅ 5/5 177.66 0.05 n/a
rms_norm_8_cols_2_channels_2048_tile_128 ✅ 5/5 231.42 0.04 n/a
rope_1c_32rows_512cols_32arows_0m ✅ 5/5 177.48 0.56 n/a
rope_1c_32rows_512cols_8arows_0m ✅ 5/5 212.16 0.35 n/a
rope_2c_32rows_512cols_32arows_0m ✅ 5/5 190.26 0.52 n/a
rope_2c_32rows_512cols_8arows_0m ✅ 5/5 196.86 0.38 n/a
rope_8c_32rows_512cols_32arows_0m ✅ 5/5 203.24 0.51 n/a
rope_8c_32rows_512cols_8arows_0m ✅ 5/5 239.26 0.32 n/a
sigmoid_1_cols_1_channels_2048_tile_2048 ✅ 5/5 152.74 0.06 n/a
sigmoid_2_cols_1_channels_2048_tile_1024 ✅ 5/5 172.44 0.05 n/a
sigmoid_4_cols_1_channels_2048_tile_512 ✅ 5/5 174.28 0.05 n/a
sigmoid_8_cols_1_channels_2048_tile_256 ✅ 5/5 202.88 0.04 n/a
silu_1_cols_1_channels_2048_tile_2048 ✅ 5/5 170.74 0.05 n/a
silu_2_cols_1_channels_2048_tile_1024 ✅ 5/5 168.68 0.05 n/a
silu_4_cols_1_channels_2048_tile_512 ✅ 5/5 195.86 0.04 n/a
silu_8_cols_1_channels_2048_tile_256 ✅ 5/5 192.66 0.04 n/a
softmax_2_cols_2_channels_32768_tile_1024 ✅ 5/5 189.12 0.71 n/a
softmax_2_cols_2_channels_32768_tile_512 ✅ 5/5 199.98 0.68 n/a
softmax_4_cols_4_channels_32768_tile_2048 ✅ 5/5 305.66 0.48 n/a
swiglu_decode_1x2048x2048 ❌ 0/5 n/a n/a n/a
swiglu_prefill_256x2048x2048 ❌ 0/5 n/a n/a n/a
tanh_1_cols_1_channels_2048_tile_2048 ✅ 5/5 161.78 0.05 n/a
tanh_2_cols_1_channels_2048_tile_1024 ✅ 5/5 199.08 0.04 n/a
tanh_4_cols_1_channels_2048_tile_512 ✅ 5/5 172.46 0.05 n/a
tanh_8_cols_1_channels_2048_tile_256 ✅ 5/5 189.90 0.04 n/a
transpose_2048_M_64_N_1_cols_1_channels_64_m_64_n_8_s ✅ 5/5 184.24 2.99 n/a
transpose_2048_M_64_N_1_cols_2_channels_64_m_64_n_8_s ✅ 5/5 162.32 3.34 n/a
weighted_rms_norm_1_cols_2_channels_2048_weights_2048 ✅ 5/5 187.34 0.07 n/a
weighted_rms_norm_2_cols_2_channels_2048_weights_1024 ✅ 5/5 158.74 0.07 n/a
weighted_rms_norm_4_cols_2_channels_2048_weights_512 ✅ 5/5 169.94 0.06 n/a
weighted_rms_norm_8_cols_2_channels_2048_weights_256 ✅ 5/5 262.02 0.04 n/a
📈 Trends (vs main branch) for Small Benchmark/Test Suite

2f954d5 (2026_03_18_18_09_41)

IRONCLAD Trends

axpy_1_cols_2_channels_2048_tile_2048_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.10 (-19.72%)0.08 (-21.50%)0.08 (-22.94%)0.06 (-17.22%)0.02 (-3.82%)196.20 (+20.81%)154.58 (+27.79%)151.90 (+29.83%)117.50 (+24.47%)32.83 (+42.05%)
130b6ea — 2025-12-05 21:15:490.13 (n/a)0.11 (n/a)0.10 (n/a)0.08 (n/a)0.02 (n/a)162.40 (n/a)120.96 (n/a)117.00 (n/a)94.40 (n/a)23.11 (n/a)

axpy_1_cols_2_channels_2048_tile_2048_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.16 (-3.31%)0.14 (-1.29%)0.15 (+16.18%)0.10 (-6.80%)0.03 (+3.89%)125.70 (+7.25%)93.56 (+2.12%)83.50 (-13.92%)75.00 (+3.45%)22.38 (+18.09%)
84d3478 — 2026-02-17 23:25:130.17 (n/a)0.14 (n/a)0.13 (n/a)0.10 (n/a)0.03 (n/a)117.20 (n/a)91.62 (n/a)97.00 (n/a)72.50 (n/a)18.95 (n/a)

axpy_2_cols_2_channels_2048_tile_1024_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.10 (-7.24%)0.07 (-21.92%)0.07 (-21.61%)0.05 (-33.79%)0.02 (+59.98%)229.60 (+51.05%)178.40 (+31.70%)182.20 (+27.59%)124.00 (+7.83%)38.34 (+155.00%)
130b6ea — 2025-12-05 21:15:490.11 (n/a)0.09 (n/a)0.09 (n/a)0.08 (n/a)0.01 (n/a)152.00 (n/a)135.46 (n/a)142.80 (n/a)115.00 (n/a)15.03 (n/a)

axpy_2_cols_2_channels_2048_tile_1024_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (-0.73%)0.11 (-7.43%)0.11 (-8.63%)0.07 (-27.71%)0.02 (+50.39%)182.70 (+38.30%)120.86 (+11.97%)110.80 (+9.38%)96.10 (+0.73%)35.16 (+122.88%)
84d3478 — 2026-02-17 23:25:130.13 (n/a)0.12 (n/a)0.12 (n/a)0.09 (n/a)0.02 (n/a)132.10 (n/a)107.94 (n/a)101.30 (n/a)95.40 (n/a)15.78 (n/a)

axpy_4_cols_2_channels_2048_tile_512_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.07 (-37.81%)0.06 (-34.41%)0.06 (-36.95%)0.06 (-19.56%)0.01 (-66.31%)207.20 (+24.30%)191.58 (+48.77%)193.10 (+58.54%)167.10 (+60.83%)15.32 (-33.24%)
130b6ea — 2025-12-05 21:15:490.12 (n/a)0.10 (n/a)0.10 (n/a)0.07 (n/a)0.02 (n/a)166.70 (n/a)128.78 (n/a)121.80 (n/a)103.90 (n/a)22.95 (n/a)

axpy_4_cols_2_channels_2048_tile_512_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (-10.91%)0.11 (-7.68%)0.12 (-0.73%)0.06 (-17.94%)0.03 (+0.99%)194.10 (+21.85%)124.52 (+10.29%)103.30 (+0.68%)97.20 (+12.24%)40.32 (+39.15%)
84d3478 — 2026-02-17 23:25:130.14 (n/a)0.11 (n/a)0.12 (n/a)0.08 (n/a)0.03 (n/a)159.30 (n/a)112.90 (n/a)102.60 (n/a)86.60 (n/a)28.97 (n/a)

axpy_8_cols_2_channels_2048_tile_256_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.08 (-20.33%)0.06 (-22.87%)0.06 (-25.59%)0.05 (-30.14%)0.01 (+10.88%)263.90 (+43.11%)201.28 (+31.40%)193.60 (+34.35%)159.50 (+25.49%)41.78 (+96.52%)
130b6ea — 2025-12-05 21:15:490.10 (n/a)0.08 (n/a)0.09 (n/a)0.07 (n/a)0.01 (n/a)184.40 (n/a)153.18 (n/a)144.10 (n/a)127.10 (n/a)21.26 (n/a)

axpy_8_cols_2_channels_2048_tile_256_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (+0.01%)0.11 (+6.10%)0.11 (+0.22%)0.08 (-0.33%)0.02 (-9.83%)156.30 (+0.32%)115.26 (-6.40%)112.70 (-0.18%)95.50 (+0.00%)24.34 (-10.22%)
84d3478 — 2026-02-17 23:25:130.13 (n/a)0.10 (n/a)0.11 (n/a)0.08 (n/a)0.02 (n/a)155.80 (n/a)123.14 (n/a)112.90 (n/a)95.50 (n/a)27.11 (n/a)

dequant_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.04 (-17.50%)0.03 (-21.53%)0.03 (-25.89%)0.02 (-25.65%)0.01 (+3.94%)215.80 (+34.54%)180.82 (+28.13%)184.10 (+34.97%)138.50 (+21.17%)28.16 (+62.15%)
130b6ea — 2025-12-05 21:15:490.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.00 (n/a)160.40 (n/a)141.12 (n/a)136.40 (n/a)114.30 (n/a)17.37 (n/a)

dequant_1_cols_1_channels_2048_tile_2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-13.00%)0.04 (-0.87%)0.04 (+1.16%)0.03 (-3.17%)0.01 (-31.33%)158.40 (+3.26%)128.12 (-0.19%)123.80 (-1.12%)110.40 (+14.88%)18.36 (-15.65%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)153.40 (n/a)128.36 (n/a)125.20 (n/a)96.10 (n/a)21.77 (n/a)

dequant_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.05 (-22.55%)0.03 (-38.88%)0.03 (-53.05%)0.02 (-45.76%)0.01 (+31.99%)246.30 (+84.36%)183.80 (+73.27%)202.70 (+113.14%)113.50 (+29.12%)55.39 (+209.44%)
130b6ea — 2025-12-05 21:15:490.06 (n/a)0.05 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)133.60 (n/a)106.08 (n/a)95.10 (n/a)87.90 (n/a)17.90 (n/a)

dequant_1_cols_2_channels_2048_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (-3.74%)0.05 (-12.63%)0.05 (-5.21%)0.03 (-33.47%)0.01 (+52.60%)162.70 (+50.37%)110.84 (+18.98%)98.80 (+5.44%)81.50 (+3.95%)31.02 (+149.23%)
84d3478 — 2026-02-17 23:25:130.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)108.20 (n/a)93.16 (n/a)93.70 (n/a)78.40 (n/a)12.45 (n/a)

dequant_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.05 (-30.03%)0.03 (-41.65%)0.03 (-44.07%)0.02 (-35.95%)0.01 (-11.13%)223.90 (+56.14%)175.54 (+73.77%)174.50 (+78.79%)112.80 (+42.97%)42.62 (+88.88%)
130b6ea — 2025-12-05 21:15:490.07 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)143.40 (n/a)101.02 (n/a)97.60 (n/a)78.90 (n/a)22.56 (n/a)

dequant_2_cols_1_channels_2048_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (+8.90%)0.05 (+1.31%)0.05 (-1.85%)0.03 (+6.35%)0.01 (+7.11%)152.30 (-5.93%)115.08 (-1.20%)100.80 (+1.92%)88.10 (-8.23%)27.20 (-4.42%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)161.90 (n/a)116.48 (n/a)98.90 (n/a)96.00 (n/a)28.46 (n/a)

dequant_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.04 (-21.42%)0.03 (-26.34%)0.03 (-38.77%)0.03 (-13.48%)0.00 (-39.15%)184.50 (+15.53%)168.70 (+33.55%)178.10 (+63.39%)135.90 (+27.25%)20.42 (-9.18%)
130b6ea — 2025-12-05 21:15:490.05 (n/a)0.04 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)159.70 (n/a)126.32 (n/a)109.00 (n/a)106.80 (n/a)22.48 (n/a)

dequant_2_cols_2_channels_2048_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.07 (+27.96%)0.05 (+1.57%)0.04 (-18.06%)0.03 (+10.23%)0.01 (+22.79%)164.70 (-9.31%)119.12 (-1.39%)120.20 (+22.03%)75.60 (-21.82%)31.61 (-13.80%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)181.60 (n/a)120.80 (n/a)98.50 (n/a)96.70 (n/a)36.67 (n/a)

dequant_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.04 (-36.04%)0.03 (-32.19%)0.03 (-29.12%)0.02 (-30.21%)0.01 (-33.57%)223.60 (+43.24%)176.78 (+46.27%)162.10 (+41.08%)138.00 (+56.29%)36.57 (+49.87%)
130b6ea — 2025-12-05 21:15:490.06 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)156.10 (n/a)120.86 (n/a)114.90 (n/a)88.30 (n/a)24.40 (n/a)

dequant_4_cols_1_channels_2048_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-2.55%)0.04 (-12.79%)0.05 (-16.25%)0.03 (-12.85%)0.01 (-1.20%)164.60 (+14.78%)122.14 (+15.14%)115.90 (+19.48%)97.50 (+2.63%)25.08 (+19.84%)
84d3478 — 2026-02-17 23:25:130.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)143.40 (n/a)106.08 (n/a)97.00 (n/a)95.00 (n/a)20.93 (n/a)

dequant_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.03 (-37.16%)0.03 (-32.33%)0.03 (-29.82%)0.02 (-26.08%)0.00 (-54.45%)227.50 (+35.34%)197.04 (+45.16%)188.40 (+42.51%)170.90 (+59.12%)22.60 (-1.32%)
130b6ea — 2025-12-05 21:15:490.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)168.10 (n/a)135.74 (n/a)132.20 (n/a)107.40 (n/a)22.91 (n/a)

dequant_4_cols_2_channels_2048_tile_256_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (+20.83%)0.05 (+8.21%)0.04 (+1.42%)0.03 (+17.39%)0.01 (+14.84%)166.40 (-14.80%)122.34 (-7.97%)127.70 (-1.39%)84.60 (-17.22%)30.47 (-19.21%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)195.30 (n/a)132.94 (n/a)129.50 (n/a)102.20 (n/a)37.71 (n/a)

dequant_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.03 (-27.06%)0.03 (-28.35%)0.03 (-32.61%)0.02 (-28.39%)0.00 (-13.07%)237.90 (+39.61%)195.96 (+39.85%)201.50 (+48.38%)164.00 (+37.12%)29.37 (+63.71%)
130b6ea — 2025-12-05 21:15:490.04 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.00 (n/a)170.40 (n/a)140.12 (n/a)135.80 (n/a)119.60 (n/a)17.94 (n/a)

dequant_8_cols_1_channels_2048_tile_256_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-25.19%)0.04 (-21.02%)0.04 (-19.70%)0.04 (-19.76%)0.00 (-37.92%)142.70 (+24.63%)126.44 (+26.24%)125.70 (+24.46%)115.50 (+33.68%)10.61 (+3.97%)
84d3478 — 2026-02-17 23:25:130.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)114.50 (n/a)100.16 (n/a)101.00 (n/a)86.40 (n/a)10.20 (n/a)

dequant_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.03 (-41.54%)0.02 (-29.92%)0.02 (-27.24%)0.02 (-16.88%)0.00 (-63.85%)238.40 (+20.28%)220.92 (+38.07%)228.40 (+37.42%)182.20 (+71.08%)23.30 (-21.97%)
130b6ea — 2025-12-05 21:15:490.05 (n/a)0.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)198.20 (n/a)160.00 (n/a)166.20 (n/a)106.50 (n/a)29.86 (n/a)

dequant_8_cols_2_channels_2048_tile_128_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.04 (-26.69%)0.04 (-11.93%)0.04 (-16.99%)0.03 (+27.28%)0.01 (-62.60%)193.10 (-21.41%)147.68 (+2.96%)136.10 (+20.44%)135.40 (+36.49%)25.41 (-58.96%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.05 (n/a)0.02 (n/a)0.01 (n/a)245.70 (n/a)143.44 (n/a)113.00 (n/a)99.20 (n/a)61.90 (n/a)

eltwise_add_1_cols_2_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.10 (-34.93%)0.08 (-40.27%)0.07 (-54.72%)0.06 (-23.59%)0.02 (-43.25%)190.70 (+30.89%)163.22 (+64.40%)184.30 (+120.98%)119.20 (+53.61%)33.64 (+17.21%)
897d04e — 2026-03-06 22:12:250.16 (n/a)0.13 (n/a)0.15 (n/a)0.08 (n/a)0.03 (n/a)145.70 (n/a)99.28 (n/a)83.40 (n/a)77.60 (n/a)28.70 (n/a)

eltwise_add_2_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.11 (-12.13%)0.07 (-36.35%)0.07 (-43.77%)0.03 (-59.36%)0.03 (+26.52%)389.10 (+146.11%)216.50 (+76.71%)186.70 (+77.81%)115.40 (+13.81%)103.64 (+276.46%)
897d04e — 2026-03-06 22:12:250.12 (n/a)0.10 (n/a)0.12 (n/a)0.08 (n/a)0.02 (n/a)158.10 (n/a)122.52 (n/a)105.00 (n/a)101.40 (n/a)27.53 (n/a)

eltwise_add_4_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.07 (-53.53%)0.07 (-43.57%)0.07 (-49.91%)0.06 (-6.27%)0.01 (-85.14%)212.00 (+6.69%)182.82 (+57.96%)172.60 (+99.54%)170.70 (+115.26%)17.56 (-65.37%)
897d04e — 2026-03-06 22:12:250.15 (n/a)0.12 (n/a)0.14 (n/a)0.06 (n/a)0.04 (n/a)198.70 (n/a)115.74 (n/a)86.50 (n/a)79.30 (n/a)50.70 (n/a)

eltwise_add_8_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.08 (-31.37%)0.06 (-37.39%)0.07 (-39.30%)0.04 (-42.49%)0.01 (-21.94%)306.10 (+73.82%)203.60 (+62.70%)185.70 (+64.77%)162.40 (+45.78%)58.43 (+105.01%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.10 (n/a)0.11 (n/a)0.07 (n/a)0.02 (n/a)176.10 (n/a)125.14 (n/a)112.70 (n/a)111.40 (n/a)28.50 (n/a)

eltwise_mul_1_cols_2_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.14 (-3.54%)0.09 (-32.18%)0.07 (-45.92%)0.06 (-14.11%)0.03 (+4.47%)193.40 (+16.44%)153.74 (+49.73%)165.00 (+84.98%)86.80 (+3.70%)42.67 (+19.94%)
897d04e — 2026-03-06 22:12:250.15 (n/a)0.13 (n/a)0.14 (n/a)0.07 (n/a)0.03 (n/a)166.10 (n/a)102.68 (n/a)89.20 (n/a)83.70 (n/a)35.57 (n/a)

eltwise_mul_2_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.09 (-43.53%)0.07 (-49.19%)0.07 (-54.42%)0.05 (-47.25%)0.01 (-47.73%)254.90 (+89.66%)188.66 (+96.36%)180.70 (+119.30%)144.20 (+76.93%)41.57 (+82.03%)
897d04e — 2026-03-06 22:12:250.15 (n/a)0.13 (n/a)0.15 (n/a)0.09 (n/a)0.03 (n/a)134.40 (n/a)96.08 (n/a)82.40 (n/a)81.50 (n/a)22.84 (n/a)

eltwise_mul_4_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.08 (-31.43%)0.07 (-35.03%)0.07 (-35.09%)0.05 (-24.22%)0.01 (-34.36%)259.20 (+31.98%)195.90 (+52.33%)177.90 (+54.16%)154.50 (+45.89%)46.62 (+21.65%)
897d04e — 2026-03-06 22:12:250.12 (n/a)0.10 (n/a)0.11 (n/a)0.06 (n/a)0.02 (n/a)196.40 (n/a)128.60 (n/a)115.40 (n/a)105.90 (n/a)38.33 (n/a)

eltwise_mul_8_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.07 (-33.13%)0.06 (-41.91%)0.06 (-41.33%)0.04 (-50.45%)0.01 (+15.42%)313.30 (+101.87%)221.76 (+79.80%)193.80 (+70.45%)167.80 (+49.55%)62.93 (+242.88%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.10 (n/a)0.11 (n/a)0.08 (n/a)0.01 (n/a)155.20 (n/a)123.34 (n/a)113.70 (n/a)112.20 (n/a)18.35 (n/a)

gelu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.08 (-28.77%)0.05 (-46.81%)0.04 (-53.63%)0.03 (-48.17%)0.02 (-6.58%)260.40 (+93.03%)186.40 (+96.46%)189.50 (+115.59%)109.00 (+40.46%)55.65 (+138.87%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.09 (n/a)0.06 (n/a)0.02 (n/a)134.90 (n/a)94.88 (n/a)87.90 (n/a)77.60 (n/a)23.30 (n/a)

gelu_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.07 (-38.38%)0.05 (-54.05%)0.04 (-60.74%)0.03 (-55.20%)0.01 (-10.27%)248.70 (+123.25%)191.02 (+125.79%)196.70 (+154.79%)118.00 (+62.31%)46.77 (+198.68%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.10 (n/a)0.11 (n/a)0.07 (n/a)0.02 (n/a)111.40 (n/a)84.60 (n/a)77.20 (n/a)72.70 (n/a)15.66 (n/a)

gelu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.07 (-37.44%)0.04 (-53.89%)0.04 (-63.95%)0.03 (-36.09%)0.01 (-39.77%)243.90 (+56.45%)204.08 (+114.60%)220.30 (+177.46%)122.50 (+59.71%)48.72 (+42.56%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.10 (n/a)0.05 (n/a)0.02 (n/a)155.90 (n/a)95.10 (n/a)79.40 (n/a)76.70 (n/a)34.18 (n/a)

gelu_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.05 (-57.55%)0.04 (-56.34%)0.04 (-62.53%)0.04 (-28.82%)0.00 (-86.08%)213.10 (+40.47%)198.10 (+113.84%)204.30 (+166.71%)176.40 (+135.51%)14.99 (-54.91%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.10 (n/a)0.11 (n/a)0.05 (n/a)0.02 (n/a)151.70 (n/a)92.64 (n/a)76.60 (n/a)74.90 (n/a)33.24 (n/a)

gelu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.05 (-57.03%)0.04 (-55.61%)0.04 (-58.36%)0.03 (-39.51%)0.01 (-76.50%)250.10 (+65.30%)203.42 (+113.77%)201.20 (+140.10%)173.80 (+132.66%)28.58 (-10.60%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.10 (n/a)0.05 (n/a)0.02 (n/a)151.30 (n/a)95.16 (n/a)83.80 (n/a)74.70 (n/a)31.97 (n/a)

gelu_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.06 (-36.75%)0.05 (-47.06%)0.04 (-53.95%)0.03 (-26.16%)0.01 (-53.80%)245.80 (+35.43%)187.24 (+79.35%)184.30 (+117.08%)131.10 (+58.14%)41.13 (-4.71%)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.09 (n/a)0.10 (n/a)0.05 (n/a)0.02 (n/a)181.50 (n/a)104.40 (n/a)84.90 (n/a)82.90 (n/a)43.16 (n/a)

gelu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.06 (-44.73%)0.04 (-52.15%)0.04 (-56.61%)0.03 (-38.99%)0.01 (-50.24%)234.80 (+63.85%)206.00 (+106.45%)210.50 (+130.56%)147.60 (+80.88%)34.41 (+37.86%)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.09 (n/a)0.09 (n/a)0.06 (n/a)0.02 (n/a)143.30 (n/a)99.78 (n/a)91.30 (n/a)81.60 (n/a)24.96 (n/a)

gelu_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.04 (-53.72%)0.04 (-52.04%)0.03 (-56.33%)0.03 (-30.96%)0.00 (-81.41%)251.50 (+44.79%)233.98 (+100.19%)243.20 (+129.00%)212.20 (+116.09%)18.02 (-43.72%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.07 (n/a)0.08 (n/a)0.05 (n/a)0.01 (n/a)173.70 (n/a)116.88 (n/a)106.20 (n/a)98.20 (n/a)32.02 (n/a)

gemm_1792x896x1152_64x32x48_8cols_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
2f954d5 — 2026-03-18 18:02:495.20 (+4.14%)4.13 (-15.37%)4.08 (-17.83%)3.29 (-27.21%)0.69 (+232.26%)2862.50 (+37.38%)2324.70 (+20.54%)2304.60 (+21.70%)1807.50 (-3.97%)375.19 (+332.37%)2046.74 (+4.14%)1625.82 (-15.37%)1605.25 (-17.83%)1292.38 (-27.21%)271.02 (+232.26%)
897d04e — 2026-03-06 22:12:255.00 (n/a)4.88 (n/a)4.97 (n/a)4.51 (n/a)0.21 (n/a)2083.60 (n/a)1928.60 (n/a)1893.70 (n/a)1882.30 (n/a)86.77 (n/a)1965.38 (n/a)1921.12 (n/a)1953.55 (n/a)1775.47 (n/a)81.57 (n/a)

gemm_192x384x64_48x96x16_4cols

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
2f954d5 — 2026-03-18 18:02:490.94 (-54.13%)0.65 (-60.54%)0.52 (-72.33%)0.48 (-38.36%)0.22 (-58.32%)456.80 (+62.22%)367.02 (+141.91%)428.70 (+261.47%)235.30 (+118.07%)108.99 (+47.47%)40.10 (-54.13%)27.93 (-60.54%)22.01 (-72.33%)20.66 (-38.36%)9.37 (-58.32%)
897d04e — 2026-03-06 22:12:252.05 (n/a)1.66 (n/a)1.86 (n/a)0.79 (n/a)0.53 (n/a)281.60 (n/a)151.72 (n/a)118.60 (n/a)107.90 (n/a)73.91 (n/a)87.43 (n/a)70.78 (n/a)79.55 (n/a)33.52 (n/a)22.49 (n/a)

gemm_192x384x64_48x96x16_4cols_bcolmaj_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
2f954d5 — 2026-03-18 18:02:491.08 (-51.79%)0.89 (-51.08%)0.97 (-47.88%)0.49 (-62.79%)0.24 (-26.97%)451.10 (+168.83%)269.60 (+115.99%)227.00 (+91.89%)205.50 (+107.37%)103.26 (+303.73%)45.92 (-51.79%)38.09 (-51.08%)41.58 (-47.88%)20.92 (-62.79%)10.20 (-26.97%)
897d04e — 2026-03-06 22:12:252.23 (n/a)1.82 (n/a)1.87 (n/a)1.32 (n/a)0.33 (n/a)167.80 (n/a)124.82 (n/a)118.30 (n/a)99.10 (n/a)25.58 (n/a)95.25 (n/a)77.86 (n/a)79.77 (n/a)56.23 (n/a)13.96 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.72 (-0.11%)3.55 (-2.75%)3.56 (-2.75%)3.36 (-5.33%)0.12 (+76.10%)2725.23 (-0.11%)2602.52 (-2.75%)2612.11 (-2.75%)2459.89 (-5.33%)86.27 (+76.10%)6984.00 (+5.63%)6608.60 (+2.91%)6577.00 (+2.83%)6304.00 (+0.11%)221.87 (+87.15%)
9e98049 — 2025-11-25 20:50:553.72 (n/a)3.65 (n/a)3.66 (n/a)3.54 (n/a)0.07 (n/a)2728.26 (n/a)2676.06 (n/a)2686.03 (n/a)2598.29 (n/a)48.99 (n/a)6612.00 (n/a)6422.00 (n/a)6396.00 (n/a)6297.00 (n/a)118.55 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_0_bcolmaj_1_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.40 (-3.14%)3.38 (-0.82%)3.38 (-0.82%)3.35 (+4.93%)0.02 (-84.51%)2493.81 (-3.14%)2475.27 (-0.82%)2479.06 (-0.82%)2453.92 (+4.93%)13.33 (-84.51%)7001.00 (-4.70%)6940.80 (+0.71%)6930.00 (+0.83%)6889.00 (+3.24%)37.43 (-84.76%)
9e98049 — 2025-11-25 20:50:553.51 (n/a)3.40 (n/a)3.41 (n/a)3.19 (n/a)0.12 (n/a)2574.53 (n/a)2495.72 (n/a)2499.62 (n/a)2338.67 (n/a)86.02 (n/a)7346.00 (n/a)6892.20 (n/a)6873.00 (n/a)6673.00 (n/a)245.57 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_1_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.02 (+0.52%)2.87 (-2.05%)2.86 (-2.24%)2.70 (-6.16%)0.11 (+131.66%)2216.19 (+0.52%)2104.44 (-2.05%)2094.08 (-2.24%)1980.84 (-6.16%)80.25 (+131.67%)8673.00 (+6.56%)8175.60 (+2.21%)8204.00 (+2.29%)7752.00 (-0.51%)314.03 (+145.36%)
9e98049 — 2025-11-25 20:50:553.01 (n/a)2.93 (n/a)2.92 (n/a)2.88 (n/a)0.05 (n/a)2204.81 (n/a)2148.41 (n/a)2142.13 (n/a)2110.81 (n/a)34.64 (n/a)8139.00 (n/a)7998.60 (n/a)8020.00 (n/a)7792.00 (n/a)127.99 (n/a)

gemm_2048x2048x2048_64x64x64_1cols

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
2f954d5 — 2026-03-18 18:02:490.53 (+0.33%)0.53 (+0.87%)0.53 (+1.06%)0.52 (+0.89%)0.00 (-73.95%)47972.90 (-0.89%)47900.00 (-0.87%)47885.90 (-1.04%)47855.10 (-0.33%)43.97 (-74.26%)359.00 (+0.33%)358.66 (+0.87%)358.77 (+1.06%)358.12 (+0.89%)0.33 (-73.95%)
897d04e — 2026-03-06 22:12:250.52 (n/a)0.52 (n/a)0.52 (n/a)0.52 (n/a)0.00 (n/a)48401.80 (n/a)48318.62 (n/a)48391.10 (n/a)48013.20 (n/a)170.80 (n/a)357.82 (n/a)355.56 (n/a)355.02 (n/a)354.94 (n/a)1.26 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:491.04 (+0.94%)1.03 (+0.23%)1.03 (+0.09%)1.02 (+0.07%)0.00 (+197.51%)24576.00 (-0.07%)24485.82 (-0.23%)24530.40 (-0.09%)24266.40 (-0.94%)111.26 (+194.35%)
0a6c11c — 2025-12-04 00:39:101.03 (n/a)1.03 (n/a)1.03 (n/a)1.02 (n/a)0.00 (n/a)24594.30 (n/a)24542.14 (n/a)24551.90 (n/a)24495.50 (n/a)37.80 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:051.04 (+0.28%)1.03 (-0.15%)1.03 (+0.05%)1.02 (-0.52%)0.01 (+42.72%)24719.80 (+0.52%)24512.42 (+0.16%)24529.20 (-0.05%)24092.60 (-0.28%)252.92 (+43.15%)713.08 (+0.28%)700.92 (-0.15%)700.38 (+0.05%)694.98 (-0.52%)7.30 (+42.71%)
50ad37c — 2025-12-29 20:35:531.04 (n/a)1.03 (n/a)1.03 (n/a)1.02 (n/a)0.01 (n/a)24592.30 (n/a)24473.70 (n/a)24540.90 (n/a)24160.80 (n/a)176.68 (n/a)711.06 (n/a)702.00 (n/a)700.05 (n/a)698.59 (n/a)5.12 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_1_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:490.99 (+1.69%)0.97 (+0.29%)0.97 (-0.04%)0.97 (-0.15%)0.01 (+434.92%)26025.30 (+0.15%)25839.34 (-0.29%)25916.70 (+0.04%)25446.00 (-1.67%)202.05 (+425.44%)
0a6c11c — 2025-12-04 00:39:100.97 (n/a)0.97 (n/a)0.97 (n/a)0.97 (n/a)0.00 (n/a)25986.90 (n/a)25913.52 (n/a)25905.70 (n/a)25877.00 (n/a)38.45 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_1_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:050.97 (-0.01%)0.97 (+0.03%)0.97 (+0.04%)0.97 (+0.00%)0.00 (-6.43%)25927.90 (-0.00%)25903.20 (-0.03%)25898.60 (-0.04%)25892.80 (+0.01%)14.05 (-6.56%)663.50 (-0.01%)663.23 (+0.03%)663.35 (+0.04%)662.60 (+0.00%)0.36 (-6.43%)
50ad37c — 2025-12-29 20:35:530.97 (n/a)0.97 (n/a)0.97 (n/a)0.97 (n/a)0.00 (n/a)25928.10 (n/a)25910.14 (n/a)25908.10 (n/a)25891.20 (n/a)15.04 (n/a)663.54 (n/a)663.06 (n/a)663.11 (n/a)662.60 (n/a)0.38 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_1_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:490.92 (-1.02%)0.92 (-0.66%)0.92 (-0.10%)0.91 (-0.78%)0.01 (-4.54%)27751.40 (+0.79%)27480.74 (+0.66%)27354.80 (+0.10%)27320.50 (+1.03%)182.69 (-2.84%)
0a6c11c — 2025-12-04 00:39:100.93 (n/a)0.92 (n/a)0.92 (n/a)0.91 (n/a)0.01 (n/a)27534.80 (n/a)27299.32 (n/a)27326.70 (n/a)27042.30 (n/a)188.04 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_1_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:050.92 (+0.11%)0.91 (-0.12%)0.91 (-0.21%)0.91 (-0.24%)0.00 (+30.34%)27723.50 (+0.24%)27548.94 (+0.12%)27522.90 (+0.21%)27403.60 (-0.11%)124.26 (+30.58%)626.92 (+0.11%)623.62 (-0.12%)624.20 (-0.21%)619.69 (-0.24%)2.81 (+30.34%)
50ad37c — 2025-12-29 20:35:530.92 (n/a)0.91 (n/a)0.92 (n/a)0.91 (n/a)0.00 (n/a)27657.10 (n/a)27515.58 (n/a)27464.70 (n/a)27434.60 (n/a)95.16 (n/a)626.21 (n/a)624.37 (n/a)625.52 (n/a)621.17 (n/a)2.16 (n/a)

gemm_2048x2048x2048_64x64x64_2cols_bcolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
2f954d5 — 2026-03-18 18:02:490.93 (+1.49%)0.92 (+2.44%)0.92 (+2.01%)0.92 (+3.95%)0.00 (-68.27%)27365.20 (-3.80%)27230.48 (-2.40%)27228.10 (-1.97%)27116.30 (-1.47%)108.02 (-69.97%)633.56 (+1.49%)630.91 (+2.44%)630.96 (+2.01%)627.80 (+3.95%)2.50 (-68.27%)
897d04e — 2026-03-06 22:12:250.91 (n/a)0.90 (n/a)0.91 (n/a)0.88 (n/a)0.01 (n/a)28447.10 (n/a)27899.18 (n/a)27774.40 (n/a)27519.50 (n/a)359.64 (n/a)624.28 (n/a)615.87 (n/a)618.55 (n/a)603.92 (n/a)7.88 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
31ab24e — 2025-11-22 02:01:563.65 (-1.06%)3.59 (+0.83%)3.63 (-0.26%)3.45 (+2.66%)0.07 (-44.71%)2678.08 (-1.06%)2632.13 (+0.83%)2658.60 (-0.26%)2528.31 (+2.66%)53.86 (-44.71%)6795.00 (-2.59%)6529.80 (-0.92%)6462.00 (+0.26%)6415.00 (+1.07%)137.14 (-45.27%)
d9a64c5 — 2025-11-21 20:49:353.69 (n/a)3.56 (n/a)3.64 (n/a)3.36 (n/a)0.13 (n/a)2706.77 (n/a)2610.50 (n/a)2665.61 (n/a)2462.71 (n/a)97.41 (n/a)6976.00 (n/a)6590.40 (n/a)6445.00 (n/a)6347.00 (n/a)250.58 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.98 (+0.20%)3.84 (-2.95%)3.76 (-5.27%)3.72 (-4.97%)0.13 (+473.97%)6760.00 (+5.23%)6559.50 (+3.13%)6697.10 (+5.56%)6321.00 (-0.20%)218.68 (+499.40%)2717.91 (+0.20%)2621.44 (-2.95%)2565.27 (-5.27%)2541.40 (-4.97%)88.37 (+473.97%)
331dcca — 2026-01-08 18:34:103.97 (n/a)3.96 (n/a)3.97 (n/a)3.92 (n/a)0.02 (n/a)6423.80 (n/a)6360.32 (n/a)6344.30 (n/a)6333.90 (n/a)36.48 (n/a)2712.39 (n/a)2701.17 (n/a)2707.91 (n/a)2674.40 (n/a)15.40 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_1_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.75 (+0.28%)3.70 (+1.25%)3.71 (-0.21%)3.65 (+4.77%)0.04 (-66.90%)6901.80 (-4.56%)6795.46 (-1.30%)6790.20 (+0.21%)6715.90 (-0.28%)67.05 (-68.40%)2558.08 (+0.28%)2528.33 (+1.25%)2530.10 (-0.21%)2489.18 (+4.77%)24.81 (-66.90%)
331dcca — 2026-01-08 18:34:103.74 (n/a)3.66 (n/a)3.71 (n/a)3.48 (n/a)0.11 (n/a)7231.20 (n/a)6885.18 (n/a)6775.70 (n/a)6734.90 (n/a)212.21 (n/a)2550.87 (n/a)2497.04 (n/a)2535.50 (n/a)2375.78 (n/a)74.96 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_1_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.62 (-0.77%)3.54 (+0.14%)3.60 (-0.04%)3.44 (+2.17%)0.09 (-25.77%)7307.40 (-2.13%)7103.46 (-0.18%)6982.50 (+0.04%)6948.20 (+0.77%)182.95 (-26.55%)2472.55 (-0.77%)2419.80 (+0.14%)2460.41 (-0.04%)2351.04 (+2.17%)61.76 (-25.77%)
331dcca — 2026-01-08 18:34:103.65 (n/a)3.54 (n/a)3.61 (n/a)3.37 (n/a)0.12 (n/a)7466.20 (n/a)7116.40 (n/a)6979.60 (n/a)6895.10 (n/a)249.09 (n/a)2491.61 (n/a)2416.45 (n/a)2461.45 (n/a)2301.00 (n/a)83.20 (n/a)

gemm_2048x2048x2048_64x64x64_8cols_bcolmaj_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
2f954d5 — 2026-03-18 18:02:493.32 (-5.30%)3.23 (-7.29%)3.21 (-7.72%)3.15 (-9.32%)0.07 (+481.75%)7978.70 (+10.27%)7785.86 (+7.91%)7831.60 (+8.36%)7572.90 (+5.59%)172.59 (+576.46%)2268.60 (-5.30%)2207.42 (-7.29%)2193.67 (-7.72%)2153.22 (-9.32%)49.14 (+481.74%)
897d04e — 2026-03-06 22:12:253.51 (n/a)3.49 (n/a)3.48 (n/a)3.48 (n/a)0.01 (n/a)7235.40 (n/a)7215.10 (n/a)7227.20 (n/a)7171.80 (n/a)25.51 (n/a)2395.46 (n/a)2381.12 (n/a)2377.13 (n/a)2374.41 (n/a)8.45 (n/a)

gemm_384x1536x1792_32x48x64_4cols_bcolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
2f954d5 — 2026-03-18 18:02:494.23 (-2.68%)3.62 (-13.03%)3.81 (-11.83%)2.94 (-16.65%)0.51 (+42.51%)2740.40 (+19.98%)2261.72 (+16.12%)2116.90 (+13.42%)1907.30 (+2.75%)332.29 (+76.73%)1108.31 (-2.68%)950.21 (-13.03%)998.59 (-11.83%)771.40 (-16.65%)133.10 (+42.51%)
897d04e — 2026-03-06 22:12:254.34 (n/a)4.17 (n/a)4.32 (n/a)3.53 (n/a)0.36 (n/a)2284.00 (n/a)1947.78 (n/a)1866.50 (n/a)1856.20 (n/a)188.02 (n/a)1138.86 (n/a)1092.52 (n/a)1132.55 (n/a)925.53 (n/a)93.39 (n/a)

gemm_896x1792x640_32x64x80_8cols_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
2f954d5 — 2026-03-18 18:02:495.01 (-25.44%)3.90 (-38.12%)3.49 (-47.18%)3.23 (-37.83%)0.79 (+21.59%)2060.20 (+60.85%)1756.22 (+64.97%)1904.00 (+89.30%)1327.00 (+34.12%)325.15 (+162.55%)1548.72 (-25.44%)1206.06 (-38.12%)1079.41 (-47.18%)997.59 (-37.83%)242.84 (+21.59%)
897d04e — 2026-03-06 22:12:256.72 (n/a)6.31 (n/a)6.61 (n/a)5.19 (n/a)0.65 (n/a)1280.80 (n/a)1064.58 (n/a)1005.80 (n/a)989.40 (n/a)123.84 (n/a)2077.17 (n/a)1949.09 (n/a)2043.44 (n/a)1604.60 (n/a)199.72 (n/a)

layer_norm_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.05 (-52.95%)0.04 (-48.58%)0.04 (-49.04%)0.03 (-36.27%)0.01 (-75.26%)234.80 (+56.85%)194.70 (+84.87%)185.70 (+96.09%)166.70 (+112.63%)26.62 (-14.42%)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)149.70 (n/a)105.32 (n/a)94.70 (n/a)78.40 (n/a)31.11 (n/a)

layer_norm_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.05 (-52.70%)0.04 (-55.98%)0.04 (-63.34%)0.03 (-43.45%)0.01 (-69.73%)247.00 (+76.81%)207.32 (+118.37%)212.20 (+172.75%)158.30 (+111.35%)31.84 (+12.82%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.11 (n/a)0.06 (n/a)0.02 (n/a)139.70 (n/a)94.94 (n/a)77.80 (n/a)74.90 (n/a)28.23 (n/a)

layer_norm_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.06 (-42.92%)0.05 (-48.57%)0.04 (-53.17%)0.04 (-41.52%)0.01 (-43.75%)232.90 (+71.00%)184.82 (+93.85%)182.40 (+113.58%)136.40 (+75.32%)38.62 (+63.42%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.10 (n/a)0.06 (n/a)0.02 (n/a)136.20 (n/a)95.34 (n/a)85.40 (n/a)77.80 (n/a)23.63 (n/a)

layer_norm_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.05 (-52.34%)0.04 (-53.65%)0.04 (-55.49%)0.02 (-42.86%)0.01 (-59.92%)376.20 (+74.98%)231.78 (+104.21%)212.40 (+124.76%)159.50 (+109.87%)83.63 (+45.82%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.08 (n/a)0.09 (n/a)0.04 (n/a)0.03 (n/a)215.00 (n/a)113.50 (n/a)94.50 (n/a)76.00 (n/a)57.35 (n/a)

layer_norm_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.06 (-44.56%)0.04 (-50.09%)0.04 (-52.39%)0.03 (-38.42%)0.01 (-42.32%)307.80 (+62.43%)212.02 (+98.52%)190.50 (+110.03%)140.70 (+80.38%)73.50 (+57.96%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.09 (n/a)0.04 (n/a)0.02 (n/a)189.50 (n/a)106.80 (n/a)90.70 (n/a)78.00 (n/a)46.53 (n/a)

layer_norm_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.06 (-39.18%)0.04 (-46.14%)0.04 (-48.92%)0.03 (-31.09%)0.01 (-44.27%)260.40 (+45.15%)199.58 (+82.00%)183.60 (+95.74%)140.40 (+64.40%)51.16 (+30.35%)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)179.40 (n/a)109.66 (n/a)93.80 (n/a)85.40 (n/a)39.25 (n/a)

layer_norm_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.05 (-49.51%)0.03 (-54.05%)0.03 (-57.51%)0.02 (-55.42%)0.01 (-43.67%)352.90 (+124.35%)249.74 (+120.97%)238.70 (+135.40%)168.10 (+98.00%)68.26 (+145.23%)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.08 (n/a)0.08 (n/a)0.05 (n/a)0.02 (n/a)157.30 (n/a)113.02 (n/a)101.40 (n/a)84.90 (n/a)27.83 (n/a)

layer_norm_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.05 (-44.41%)0.03 (-55.66%)0.03 (-59.51%)0.02 (-51.61%)0.01 (-44.95%)409.10 (+106.62%)275.98 (+127.03%)243.90 (+146.86%)173.80 (+79.92%)89.05 (+104.70%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)198.00 (n/a)121.56 (n/a)98.80 (n/a)96.60 (n/a)43.50 (n/a)

matrix_vector_mul_128x128_32_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:490.37 (+17.13%)0.30 (+24.87%)0.30 (+38.03%)0.23 (+12.47%)0.06 (+35.23%)
0a6c11c — 2025-12-04 00:39:100.32 (n/a)0.24 (n/a)0.22 (n/a)0.20 (n/a)0.04 (n/a)

matrix_vector_mul_128x128_32_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:100.26 (-1.81%)0.24 (+4.06%)0.24 (+0.52%)0.21 (+16.97%)0.02 (-48.16%)0.26 (-1.81%)0.24 (+4.06%)0.24 (+0.52%)0.20 (+16.97%)0.02 (-48.16%)
a4b6ffe — 2026-01-05 19:24:050.26 (n/a)0.23 (n/a)0.24 (n/a)0.18 (n/a)0.04 (n/a)0.26 (n/a)0.23 (n/a)0.24 (n/a)0.17 (n/a)0.04 (n/a)

matrix_vector_mul_128x128_32tsi_128tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
2f954d5 — 2026-03-18 18:02:490.24 (n/a)0.19 (n/a)0.19 (n/a)0.13 (n/a)0.05 (n/a)0.24 (n/a)0.18 (n/a)0.18 (n/a)0.13 (n/a)0.05 (n/a)

matrix_vector_mul_128x128_32tsi_128tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:250.37 (+5.44%)0.31 (-0.05%)0.34 (-1.99%)0.22 (-7.87%)0.06 (+15.13%)0.36 (+5.44%)0.31 (-0.05%)0.34 (-1.99%)0.21 (-7.87%)0.06 (+15.13%)
84d3478 — 2026-02-17 23:25:130.35 (n/a)0.31 (n/a)0.35 (n/a)0.23 (n/a)0.05 (n/a)0.35 (n/a)0.31 (n/a)0.34 (n/a)0.23 (n/a)0.05 (n/a)

matrix_vector_mul_2048x8192_1_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4913.57 (+3.47%)13.15 (+1.09%)13.10 (+0.16%)12.52 (-1.58%)0.37 (+153.19%)
0a6c11c — 2025-12-04 00:39:1013.11 (n/a)13.01 (n/a)13.07 (n/a)12.72 (n/a)0.15 (n/a)

matrix_vector_mul_2048x8192_1_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1013.43 (-0.07%)13.30 (+4.62%)13.27 (+1.29%)13.16 (+16.70%)0.12 (-86.69%)13.42 (-0.07%)13.29 (+4.62%)13.26 (+1.29%)13.16 (+16.70%)0.12 (-86.69%)
a4b6ffe — 2026-01-05 19:24:0513.44 (n/a)12.71 (n/a)13.10 (n/a)11.28 (n/a)0.87 (n/a)13.43 (n/a)12.70 (n/a)13.09 (n/a)11.27 (n/a)0.87 (n/a)

matrix_vector_mul_2048x8192_1_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4925.06 (+2.47%)23.71 (+1.81%)24.13 (+0.50%)20.90 (+5.41%)1.48 (-15.60%)
0a6c11c — 2025-12-04 00:39:1024.46 (n/a)23.29 (n/a)24.01 (n/a)19.83 (n/a)1.75 (n/a)

matrix_vector_mul_2048x8192_1_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1024.46 (-0.60%)21.07 (-8.03%)19.75 (-17.83%)18.08 (+1.18%)3.03 (+7.07%)24.44 (-0.60%)21.06 (-8.03%)19.74 (-17.83%)18.07 (+1.18%)3.03 (+7.07%)
a4b6ffe — 2026-01-05 19:24:0524.60 (n/a)22.91 (n/a)24.04 (n/a)17.87 (n/a)2.83 (n/a)24.59 (n/a)22.90 (n/a)24.02 (n/a)17.86 (n/a)2.83 (n/a)

matrix_vector_mul_2048x8192_1_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4942.05 (+1.54%)40.50 (+12.60%)40.49 (+2.36%)38.98 (+97.53%)0.97 (-88.09%)
0a6c11c — 2025-12-04 00:39:1041.41 (n/a)35.97 (n/a)39.55 (n/a)19.73 (n/a)8.15 (n/a)

matrix_vector_mul_2048x8192_1_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1042.58 (+1.51%)41.37 (+14.29%)41.25 (+4.89%)40.33 (+91.42%)0.92 (-89.18%)42.56 (+1.51%)41.34 (+14.29%)41.23 (+4.89%)40.30 (+91.42%)0.92 (-89.18%)
a4b6ffe — 2026-01-05 19:24:0541.95 (n/a)36.20 (n/a)39.33 (n/a)21.07 (n/a)8.54 (n/a)41.92 (n/a)36.17 (n/a)39.31 (n/a)21.05 (n/a)8.53 (n/a)

matrix_vector_mul_2048x8192_1_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4946.62 (+3.11%)44.82 (+0.17%)44.78 (+0.17%)42.60 (-3.80%)1.60 (+367.72%)
0a6c11c — 2025-12-04 00:39:1045.21 (n/a)44.75 (n/a)44.70 (n/a)44.29 (n/a)0.34 (n/a)

matrix_vector_mul_2048x8192_1_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1046.04 (+4.14%)45.06 (+3.47%)44.55 (+2.76%)44.30 (+2.84%)0.87 (+66.58%)46.01 (+4.14%)45.03 (+3.47%)44.52 (+2.76%)44.27 (+2.84%)0.87 (+66.58%)
a4b6ffe — 2026-01-05 19:24:0544.21 (n/a)43.55 (n/a)43.35 (n/a)43.07 (n/a)0.52 (n/a)44.18 (n/a)43.52 (n/a)43.32 (n/a)43.05 (n/a)0.52 (n/a)

matrix_vector_mul_2048x8192_1tsi_1024tso_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
2f954d5 — 2026-03-18 18:02:4924.09 (n/a)23.45 (n/a)23.76 (n/a)22.11 (n/a)0.81 (n/a)24.08 (n/a)23.43 (n/a)23.75 (n/a)22.10 (n/a)0.81 (n/a)

matrix_vector_mul_2048x8192_1tsi_1024tso_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2524.47 (-3.32%)24.16 (-3.13%)24.20 (-3.09%)23.84 (-3.04%)0.25 (-14.36%)24.46 (-3.32%)24.15 (-3.13%)24.18 (-3.09%)23.83 (-3.04%)0.25 (-14.36%)
84d3478 — 2026-02-17 23:25:1325.31 (n/a)24.94 (n/a)24.97 (n/a)24.59 (n/a)0.29 (n/a)25.30 (n/a)24.93 (n/a)24.95 (n/a)24.58 (n/a)0.29 (n/a)

matrix_vector_mul_2048x8192_1tsi_2048tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
2f954d5 — 2026-03-18 18:02:4913.14 (n/a)12.96 (n/a)13.03 (n/a)12.51 (n/a)0.26 (n/a)13.13 (n/a)12.95 (n/a)13.02 (n/a)12.50 (n/a)0.26 (n/a)

matrix_vector_mul_2048x8192_1tsi_2048tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2513.25 (-3.04%)13.02 (-2.92%)13.10 (-2.64%)12.56 (-4.14%)0.27 (+32.55%)13.24 (-3.04%)13.01 (-2.92%)13.09 (-2.64%)12.55 (-4.14%)0.27 (+32.55%)
84d3478 — 2026-02-17 23:25:1313.66 (n/a)13.41 (n/a)13.45 (n/a)13.10 (n/a)0.20 (n/a)13.65 (n/a)13.40 (n/a)13.44 (n/a)13.09 (n/a)0.20 (n/a)

matrix_vector_mul_2048x8192_1tsi_256tso_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
2f954d5 — 2026-03-18 18:02:4944.66 (n/a)41.35 (n/a)40.89 (n/a)39.61 (n/a)1.93 (n/a)44.64 (n/a)41.33 (n/a)40.86 (n/a)39.59 (n/a)1.93 (n/a)

matrix_vector_mul_2048x8192_1tsi_256tso_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2547.71 (+3.51%)45.38 (+9.55%)45.22 (-1.27%)43.47 (+75.24%)1.52 (-83.67%)47.68 (+3.51%)45.35 (+9.55%)45.19 (-1.27%)43.45 (+75.24%)1.52 (-83.67%)
84d3478 — 2026-02-17 23:25:1346.09 (n/a)41.43 (n/a)45.80 (n/a)24.81 (n/a)9.31 (n/a)46.06 (n/a)41.40 (n/a)45.77 (n/a)24.79 (n/a)9.30 (n/a)

matrix_vector_mul_2048x8192_1tsi_512tso_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
2f954d5 — 2026-03-18 18:02:4940.50 (n/a)38.87 (n/a)38.83 (n/a)37.48 (n/a)1.10 (n/a)40.48 (n/a)38.84 (n/a)38.80 (n/a)37.46 (n/a)1.10 (n/a)

matrix_vector_mul_2048x8192_1tsi_512tso_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2540.84 (-0.32%)39.53 (-1.25%)39.40 (-1.97%)38.87 (-0.69%)0.77 (-6.24%)40.81 (-0.32%)39.50 (-1.25%)39.37 (-1.97%)38.84 (-0.69%)0.77 (-6.24%)
84d3478 — 2026-02-17 23:25:1340.97 (n/a)40.03 (n/a)40.19 (n/a)39.14 (n/a)0.82 (n/a)40.95 (n/a)40.00 (n/a)40.17 (n/a)39.11 (n/a)0.82 (n/a)

matrix_vector_mul_8192x2048_4_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4913.29 (+1.63%)13.19 (+6.08%)13.19 (+1.17%)13.09 (+26.86%)0.08 (-92.94%)
0a6c11c — 2025-12-04 00:39:1013.08 (n/a)12.44 (n/a)13.03 (n/a)10.32 (n/a)1.07 (n/a)

matrix_vector_mul_8192x2048_4_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1013.68 (+3.42%)13.30 (+8.54%)13.38 (+7.25%)12.72 (+14.96%)0.35 (-66.09%)13.67 (+3.42%)13.29 (+8.54%)13.38 (+7.25%)12.71 (+14.96%)0.35 (-66.09%)
a4b6ffe — 2026-01-05 19:24:0513.23 (n/a)12.25 (n/a)12.48 (n/a)11.07 (n/a)1.04 (n/a)13.22 (n/a)12.24 (n/a)12.47 (n/a)11.06 (n/a)1.04 (n/a)

matrix_vector_mul_8192x2048_4_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4925.58 (+4.24%)24.23 (+6.59%)24.03 (+0.29%)23.58 (+37.29%)0.70 (-74.97%)
0a6c11c — 2025-12-04 00:39:1024.54 (n/a)22.73 (n/a)23.96 (n/a)17.18 (n/a)2.79 (n/a)

matrix_vector_mul_8192x2048_4_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1025.08 (+1.32%)24.68 (+13.42%)24.61 (+3.26%)24.47 (+48.32%)0.24 (-93.56%)25.07 (+1.32%)24.67 (+13.42%)24.59 (+3.26%)24.46 (+48.32%)0.24 (-93.56%)
a4b6ffe — 2026-01-05 19:24:0524.76 (n/a)21.76 (n/a)23.83 (n/a)16.50 (n/a)3.74 (n/a)24.74 (n/a)21.75 (n/a)23.82 (n/a)16.49 (n/a)3.73 (n/a)

matrix_vector_mul_8192x2048_4_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4940.94 (-2.19%)39.74 (-1.10%)39.76 (-0.60%)38.17 (-2.14%)1.04 (+4.39%)
0a6c11c — 2025-12-04 00:39:1041.86 (n/a)40.18 (n/a)40.00 (n/a)39.00 (n/a)1.00 (n/a)

matrix_vector_mul_8192x2048_4_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1040.74 (-1.09%)37.16 (-7.15%)40.25 (+1.47%)24.51 (-37.44%)7.07 (+736.13%)40.72 (-1.09%)37.13 (-7.15%)40.23 (+1.47%)24.50 (-37.44%)7.07 (+736.13%)
a4b6ffe — 2026-01-05 19:24:0541.19 (n/a)40.02 (n/a)39.67 (n/a)39.18 (n/a)0.85 (n/a)41.17 (n/a)39.99 (n/a)39.64 (n/a)39.16 (n/a)0.85 (n/a)

matrix_vector_mul_8192x2048_4_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4946.71 (-1.04%)44.91 (-2.34%)44.71 (-2.98%)43.80 (-1.44%)0.96 (+6.93%)
0a6c11c — 2025-12-04 00:39:1047.20 (n/a)45.98 (n/a)46.08 (n/a)44.44 (n/a)0.90 (n/a)

matrix_vector_mul_8192x2048_4_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1046.28 (+5.68%)44.89 (+14.59%)44.74 (+4.26%)44.20 (+62.80%)0.84 (-87.96%)46.25 (+5.68%)44.86 (+14.59%)44.71 (+4.26%)44.17 (+62.80%)0.84 (-87.96%)
a4b6ffe — 2026-01-05 19:24:0543.79 (n/a)39.17 (n/a)42.91 (n/a)27.15 (n/a)6.98 (n/a)43.77 (n/a)39.15 (n/a)42.88 (n/a)27.13 (n/a)6.98 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
2f954d5 — 2026-03-18 18:02:4913.12 (n/a)12.61 (n/a)13.04 (n/a)11.84 (n/a)0.65 (n/a)13.11 (n/a)12.60 (n/a)13.03 (n/a)11.83 (n/a)0.65 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2513.34 (-0.79%)13.20 (-0.36%)13.18 (-1.55%)13.12 (+1.75%)0.08 (-65.93%)13.33 (-0.79%)13.19 (-0.36%)13.17 (-1.55%)13.11 (+1.75%)0.08 (-65.93%)
84d3478 — 2026-02-17 23:25:1313.45 (n/a)13.25 (n/a)13.39 (n/a)12.90 (n/a)0.25 (n/a)13.44 (n/a)13.24 (n/a)13.38 (n/a)12.89 (n/a)0.25 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
2f954d5 — 2026-03-18 18:02:4924.34 (n/a)23.55 (n/a)23.30 (n/a)23.20 (n/a)0.48 (n/a)24.32 (n/a)23.53 (n/a)23.29 (n/a)23.19 (n/a)0.48 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2524.10 (-5.45%)23.07 (-1.85%)24.01 (-2.88%)19.26 (+7.47%)2.13 (-32.15%)24.08 (-5.45%)23.05 (-1.85%)23.99 (-2.88%)19.25 (+7.47%)2.13 (-32.15%)
84d3478 — 2026-02-17 23:25:1325.48 (n/a)23.50 (n/a)24.72 (n/a)17.92 (n/a)3.14 (n/a)25.47 (n/a)23.48 (n/a)24.71 (n/a)17.91 (n/a)3.14 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
2f954d5 — 2026-03-18 18:02:4939.92 (n/a)38.47 (n/a)38.01 (n/a)37.69 (n/a)0.98 (n/a)39.89 (n/a)38.45 (n/a)37.99 (n/a)37.67 (n/a)0.98 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2540.80 (-0.16%)39.58 (-1.93%)39.75 (-1.71%)38.28 (-2.50%)1.10 (+67.33%)40.78 (-0.16%)39.55 (-1.93%)39.73 (-1.71%)38.26 (-2.50%)1.10 (+67.33%)
84d3478 — 2026-02-17 23:25:1340.87 (n/a)40.36 (n/a)40.45 (n/a)39.26 (n/a)0.66 (n/a)40.84 (n/a)40.33 (n/a)40.42 (n/a)39.24 (n/a)0.65 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
2f954d5 — 2026-03-18 18:02:4943.49 (n/a)40.86 (n/a)41.01 (n/a)38.62 (n/a)1.84 (n/a)43.46 (n/a)40.84 (n/a)40.98 (n/a)38.60 (n/a)1.84 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2549.31 (+6.27%)46.12 (+1.67%)45.30 (-0.27%)44.62 (+1.87%)1.85 (+85.10%)49.28 (+6.27%)46.10 (+1.67%)45.28 (-0.27%)44.59 (+1.87%)1.85 (+85.10%)
84d3478 — 2026-02-17 23:25:1346.40 (n/a)45.37 (n/a)45.43 (n/a)43.80 (n/a)1.00 (n/a)46.37 (n/a)45.34 (n/a)45.40 (n/a)43.78 (n/a)1.00 (n/a)

mem_copy_16_cores_2_chans_2048_tile_128_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.05 (-31.63%)0.03 (-42.10%)0.03 (-47.00%)0.03 (-36.14%)0.01 (-13.94%)271.30 (+56.55%)239.32 (+73.50%)252.00 (+88.62%)180.50 (+46.27%)34.71 (+88.59%)
130b6ea — 2025-12-05 21:15:490.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)173.30 (n/a)137.94 (n/a)133.60 (n/a)123.40 (n/a)18.40 (n/a)

mem_copy_16_cores_2_chans_2048_tile_128_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-2.12%)0.06 (-16.44%)0.06 (-21.48%)0.03 (-23.84%)0.02 (+10.15%)247.60 (+31.28%)148.38 (+24.08%)128.30 (+27.41%)100.10 (+2.14%)58.70 (+50.73%)
84d3478 — 2026-02-17 23:25:130.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)188.60 (n/a)119.58 (n/a)100.70 (n/a)98.00 (n/a)38.94 (n/a)

mem_copy_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (-16.99%)0.05 (-12.47%)0.05 (-7.32%)0.04 (-11.17%)0.00 (-29.61%)197.00 (+12.57%)171.00 (+13.70%)164.00 (+7.89%)153.00 (+20.47%)17.49 (-4.32%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)175.00 (n/a)150.40 (n/a)152.00 (n/a)127.00 (n/a)18.28 (n/a)

mem_copy_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (+5.56%)0.05 (-1.59%)0.05 (-1.39%)0.03 (-14.07%)0.01 (+25.03%)263.00 (+16.37%)166.00 (+4.53%)144.00 (+1.41%)126.00 (-5.26%)49.38 (+43.17%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)226.00 (n/a)158.80 (n/a)142.00 (n/a)133.00 (n/a)34.49 (n/a)

mem_copy_1_cores_1_chans_2048_tile_2048_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.06 (-27.19%)0.04 (-35.39%)0.05 (-29.96%)0.03 (-44.04%)0.01 (+13.96%)268.50 (+78.76%)195.22 (+60.44%)179.30 (+42.75%)134.50 (+37.39%)53.36 (+187.52%)
130b6ea — 2025-12-05 21:15:490.08 (n/a)0.07 (n/a)0.07 (n/a)0.05 (n/a)0.01 (n/a)150.20 (n/a)121.68 (n/a)125.60 (n/a)97.90 (n/a)18.56 (n/a)

mem_copy_1_cores_1_chans_2048_tile_2048_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (-0.24%)0.08 (-17.53%)0.08 (-27.05%)0.06 (-15.43%)0.02 (+4.62%)148.50 (+18.23%)113.02 (+23.38%)103.60 (+37.04%)74.10 (+0.14%)31.69 (+33.33%)
84d3478 — 2026-02-17 23:25:130.11 (n/a)0.09 (n/a)0.11 (n/a)0.07 (n/a)0.02 (n/a)125.60 (n/a)91.60 (n/a)75.60 (n/a)74.00 (n/a)23.77 (n/a)

mem_copy_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (-15.32%)0.05 (-17.09%)0.05 (-23.81%)0.04 (-15.54%)0.01 (-18.18%)193.00 (+18.40%)160.80 (+20.36%)168.00 (+31.25%)124.00 (+18.10%)22.92 (+11.39%)
0999a9a — 2025-11-13 19:24:080.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)163.00 (n/a)133.60 (n/a)128.00 (n/a)105.00 (n/a)20.58 (n/a)

mem_copy_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (-23.02%)0.05 (-6.98%)0.05 (-2.13%)0.04 (-15.31%)0.01 (-33.29%)196.00 (+18.07%)147.00 (+6.52%)141.00 (+2.17%)126.00 (+29.90%)25.27 (+8.13%)
0999a9a — 2025-11-13 19:24:080.08 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)166.00 (n/a)138.00 (n/a)138.00 (n/a)97.00 (n/a)23.37 (n/a)

mem_copy_2_cores_1_chans_2048_tile_1024_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.07 (-33.27%)0.05 (-45.21%)0.04 (-46.80%)0.03 (-55.73%)0.01 (+34.13%)248.20 (+125.84%)187.68 (+90.93%)189.50 (+88.00%)119.00 (+49.87%)46.56 (+341.64%)
130b6ea — 2025-12-05 21:15:490.10 (n/a)0.08 (n/a)0.08 (n/a)0.07 (n/a)0.01 (n/a)109.90 (n/a)98.30 (n/a)100.80 (n/a)79.40 (n/a)10.54 (n/a)

mem_copy_2_cores_1_chans_2048_tile_1024_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (+27.03%)0.09 (+25.79%)0.11 (+27.09%)0.05 (+12.74%)0.03 (+43.93%)161.90 (-11.29%)94.34 (-18.35%)76.30 (-21.34%)74.00 (-21.36%)38.01 (+0.49%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)182.50 (n/a)115.54 (n/a)97.00 (n/a)94.10 (n/a)37.82 (n/a)

mem_copy_2_cores_2_chans_2048_tile_1024_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.05 (-51.00%)0.04 (-46.65%)0.04 (-45.70%)0.03 (-48.98%)0.01 (-45.86%)294.20 (+96.00%)214.68 (+87.33%)205.80 (+84.24%)171.20 (+104.05%)47.53 (+123.85%)
130b6ea — 2025-12-05 21:15:490.10 (n/a)0.07 (n/a)0.07 (n/a)0.05 (n/a)0.01 (n/a)150.10 (n/a)114.60 (n/a)111.70 (n/a)83.90 (n/a)21.23 (n/a)

mem_copy_2_cores_2_chans_2048_tile_1024_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.10 (+2.70%)0.08 (-5.38%)0.09 (+2.10%)0.04 (-43.30%)0.02 (+170.50%)201.60 (+76.38%)112.08 (+16.53%)90.80 (-2.05%)84.60 (-2.65%)50.14 (+375.75%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.09 (n/a)0.09 (n/a)0.07 (n/a)0.01 (n/a)114.30 (n/a)96.18 (n/a)92.70 (n/a)86.90 (n/a)10.54 (n/a)

mem_copy_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.07 (+11.97%)0.05 (+3.09%)0.05 (+2.10%)0.03 (-11.35%)0.01 (+41.65%)229.00 (+12.81%)156.00 (-0.51%)143.00 (-2.05%)117.00 (-10.69%)38.21 (+48.71%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)203.00 (n/a)156.80 (n/a)146.00 (n/a)131.00 (n/a)25.69 (n/a)

mem_copy_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (+9.02%)0.05 (+5.59%)0.05 (+0.00%)0.04 (+3.63%)0.01 (+25.48%)193.00 (-3.50%)147.40 (-4.66%)147.00 (+0.00%)122.00 (-8.27%)25.24 (+7.48%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)200.00 (n/a)154.60 (n/a)147.00 (n/a)133.00 (n/a)23.48 (n/a)

mem_copy_4_cores_1_chans_2048_tile_512_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.05 (-43.25%)0.04 (-43.15%)0.04 (-40.59%)0.03 (-48.57%)0.01 (-33.14%)252.20 (+94.45%)200.08 (+76.59%)196.70 (+68.26%)159.30 (+76.22%)34.41 (+128.81%)
130b6ea — 2025-12-05 21:15:490.09 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)129.70 (n/a)113.30 (n/a)116.90 (n/a)90.40 (n/a)15.04 (n/a)

mem_copy_4_cores_1_chans_2048_tile_512_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.09 (-5.87%)0.08 (-3.61%)0.08 (-0.79%)0.06 (-12.74%)0.01 (+22.04%)146.30 (+14.66%)108.02 (+4.96%)99.30 (+0.81%)95.10 (+6.26%)21.49 (+48.38%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)127.60 (n/a)102.92 (n/a)98.50 (n/a)89.50 (n/a)14.49 (n/a)

mem_copy_4_cores_2_chans_2048_tile_512_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.05 (-42.04%)0.04 (-42.62%)0.04 (-43.93%)0.04 (-38.22%)0.00 (-54.01%)217.70 (+61.86%)201.32 (+72.60%)208.40 (+78.42%)169.00 (+72.45%)20.08 (+27.77%)
130b6ea — 2025-12-05 21:15:490.08 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)134.50 (n/a)116.64 (n/a)116.80 (n/a)98.00 (n/a)15.71 (n/a)

mem_copy_4_cores_2_chans_2048_tile_512_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (+13.04%)0.09 (+17.19%)0.10 (+20.43%)0.06 (+17.23%)0.02 (+16.34%)142.80 (-14.70%)91.94 (-14.71%)79.00 (-17.02%)77.70 (-11.50%)28.50 (-14.82%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)167.40 (n/a)107.80 (n/a)95.20 (n/a)87.80 (n/a)33.46 (n/a)

mem_copy_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (-13.48%)0.05 (-17.79%)0.05 (-15.33%)0.03 (-25.09%)0.01 (+6.67%)267.00 (+33.50%)176.80 (+24.51%)150.00 (+18.11%)141.00 (+15.57%)47.41 (+61.41%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.06 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)200.00 (n/a)142.00 (n/a)127.00 (n/a)122.00 (n/a)29.37 (n/a)

mem_copy_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (+13.38%)0.05 (+23.85%)0.05 (+37.91%)0.05 (+30.72%)0.00 (-56.85%)166.00 (-23.50%)153.40 (-20.60%)153.00 (-27.49%)142.00 (-11.80%)7.63 (-70.60%)
0999a9a — 2025-11-13 19:24:080.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)217.00 (n/a)193.20 (n/a)211.00 (n/a)161.00 (n/a)25.96 (n/a)

mem_copy_8_cores_1_chans_2048_tile_256_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.04 (-49.74%)0.04 (-46.88%)0.04 (-42.66%)0.03 (-50.11%)0.00 (-33.86%)235.10 (+100.43%)203.44 (+88.72%)189.30 (+74.47%)187.20 (+98.94%)21.62 (+160.70%)
130b6ea — 2025-12-05 21:15:490.09 (n/a)0.08 (n/a)0.08 (n/a)0.07 (n/a)0.01 (n/a)117.30 (n/a)107.80 (n/a)108.50 (n/a)94.10 (n/a)8.29 (n/a)

mem_copy_8_cores_1_chans_2048_tile_256_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-0.39%)0.07 (+10.07%)0.08 (+1.07%)0.06 (+38.38%)0.01 (-47.35%)136.70 (-27.75%)110.64 (-13.07%)104.70 (-1.13%)103.00 (+0.39%)14.59 (-60.78%)
84d3478 — 2026-02-17 23:25:130.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)189.20 (n/a)127.28 (n/a)105.90 (n/a)102.60 (n/a)37.20 (n/a)

mem_copy_8_cores_2_chans_2048_tile_256_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.05 (-49.00%)0.04 (-44.48%)0.04 (-39.33%)0.03 (-56.97%)0.01 (-29.58%)317.30 (+132.45%)221.76 (+83.49%)206.20 (+64.83%)179.50 (+95.96%)54.53 (+246.82%)
130b6ea — 2025-12-05 21:15:490.09 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)136.50 (n/a)120.86 (n/a)125.10 (n/a)91.60 (n/a)15.72 (n/a)

mem_copy_8_cores_2_chans_2048_tile_256_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-17.85%)0.07 (-24.66%)0.08 (-18.40%)0.04 (-32.08%)0.02 (+21.78%)185.30 (+47.18%)127.72 (+37.87%)102.00 (+22.45%)100.40 (+21.70%)38.66 (+106.34%)
84d3478 — 2026-02-17 23:25:130.10 (n/a)0.09 (n/a)0.10 (n/a)0.07 (n/a)0.01 (n/a)125.90 (n/a)92.64 (n/a)83.30 (n/a)82.50 (n/a)18.74 (n/a)

mha

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.18 (-0.55%)0.18 (-0.58%)0.18 (-0.52%)0.18 (-0.80%)0.00 (+189.71%)47636.70 (+0.81%)47490.22 (+0.59%)47455.80 (+0.52%)47430.50 (+0.56%)84.33 (+193.78%)
130b6ea — 2025-12-05 21:15:490.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47253.90 (n/a)47213.70 (n/a)47210.90 (n/a)47167.70 (n/a)28.71 (n/a)

mha0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
84d3478 — 2026-02-17 23:25:130.18 (-0.51%)0.18 (-0.36%)0.18 (-0.41%)0.18 (-0.21%)0.00 (-35.21%)47468.60 (+0.21%)47403.94 (+0.36%)47426.00 (+0.41%)47314.40 (+0.51%)68.83 (-34.73%)
1531781 — 2026-01-23 18:49:240.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47367.50 (n/a)47232.68 (n/a)47232.00 (n/a)47073.00 (n/a)105.45 (n/a)

mha_16384_64_1_8_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47533.70 (n/a)47380.56 (n/a)47337.90 (n/a)47268.00 (n/a)110.60 (n/a)

relu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.06 (-35.99%)0.05 (-39.39%)0.04 (-49.21%)0.04 (-27.93%)0.01 (-61.37%)209.60 (+38.81%)183.78 (+58.92%)183.80 (+96.79%)143.80 (+56.30%)26.43 (-16.25%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.07 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)151.00 (n/a)115.64 (n/a)93.40 (n/a)92.00 (n/a)31.56 (n/a)

relu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.06 (-37.09%)0.05 (-43.94%)0.05 (-47.99%)0.04 (-41.50%)0.01 (-37.90%)223.40 (+70.93%)185.02 (+78.52%)181.20 (+92.36%)148.30 (+58.95%)27.07 (+68.55%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.09 (n/a)0.06 (n/a)0.01 (n/a)130.70 (n/a)103.64 (n/a)94.20 (n/a)93.30 (n/a)16.06 (n/a)

relu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.06 (-48.86%)0.04 (-54.95%)0.04 (-59.03%)0.03 (-30.88%)0.01 (-64.56%)249.90 (+44.70%)202.80 (+108.21%)191.20 (+144.19%)147.80 (+95.50%)41.99 (-0.45%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.10 (n/a)0.05 (n/a)0.03 (n/a)172.70 (n/a)97.40 (n/a)78.30 (n/a)75.60 (n/a)42.18 (n/a)

relu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.06 (-31.34%)0.05 (-37.05%)0.04 (-46.10%)0.04 (-4.77%)0.01 (-62.99%)212.70 (+5.04%)184.50 (+49.95%)191.70 (+85.58%)147.60 (+45.71%)24.27 (-45.39%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)202.50 (n/a)123.04 (n/a)103.30 (n/a)101.30 (n/a)44.44 (n/a)

rms_norm_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.05 (-40.51%)0.05 (-40.40%)0.05 (-45.99%)0.04 (-29.40%)0.00 (-62.55%)202.40 (+41.64%)180.46 (+65.20%)180.20 (+85.20%)161.60 (+67.98%)17.99 (-10.65%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)142.90 (n/a)109.24 (n/a)97.30 (n/a)96.20 (n/a)20.13 (n/a)

rms_norm_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.07 (-21.09%)0.05 (-39.30%)0.04 (-47.51%)0.04 (-33.86%)0.01 (-7.85%)222.00 (+51.23%)180.46 (+66.91%)187.20 (+90.63%)122.80 (+26.73%)36.04 (+66.11%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)146.80 (n/a)108.12 (n/a)98.20 (n/a)96.90 (n/a)21.70 (n/a)

rms_norm_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.08 (-36.54%)0.06 (-45.07%)0.05 (-51.51%)0.04 (-36.52%)0.01 (-35.91%)197.20 (+57.51%)154.88 (+81.49%)160.90 (+106.28%)107.30 (+57.56%)33.12 (+46.05%)
897d04e — 2026-03-06 22:12:250.12 (n/a)0.10 (n/a)0.10 (n/a)0.07 (n/a)0.02 (n/a)125.20 (n/a)85.34 (n/a)78.00 (n/a)68.10 (n/a)22.68 (n/a)

rms_norm_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.06 (-32.94%)0.05 (-35.62%)0.05 (-39.51%)0.05 (-23.02%)0.00 (-61.37%)181.90 (+29.93%)162.22 (+53.04%)161.20 (+65.33%)144.20 (+49.12%)14.17 (-25.49%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)140.00 (n/a)106.00 (n/a)97.50 (n/a)96.70 (n/a)19.02 (n/a)

rms_norm_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.06 (-26.27%)0.05 (-34.40%)0.06 (-34.90%)0.03 (-39.39%)0.01 (-14.20%)251.10 (+64.98%)165.72 (+55.96%)146.90 (+53.66%)126.60 (+35.69%)50.17 (+95.18%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.01 (n/a)152.20 (n/a)106.26 (n/a)95.60 (n/a)93.30 (n/a)25.71 (n/a)

rms_norm_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.05 (-36.41%)0.05 (-43.79%)0.04 (-49.07%)0.04 (-39.08%)0.01 (-36.61%)214.10 (+64.19%)184.16 (+77.97%)191.50 (+96.41%)150.40 (+57.16%)24.22 (+60.64%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)130.40 (n/a)103.48 (n/a)97.50 (n/a)95.70 (n/a)15.08 (n/a)

rms_norm_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.06 (-34.53%)0.05 (-35.73%)0.05 (-40.50%)0.04 (-12.85%)0.01 (-58.35%)218.40 (+14.77%)177.66 (+48.74%)174.50 (+68.11%)145.40 (+52.73%)27.77 (-30.22%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)190.30 (n/a)119.44 (n/a)103.80 (n/a)95.20 (n/a)39.79 (n/a)

rms_norm_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.04 (-61.87%)0.04 (-59.62%)0.04 (-62.54%)0.03 (-24.79%)0.00 (-86.13%)261.80 (+32.96%)231.42 (+120.61%)221.70 (+166.79%)204.10 (+162.34%)24.09 (-53.22%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.10 (n/a)0.04 (n/a)0.03 (n/a)196.90 (n/a)104.90 (n/a)83.10 (n/a)77.80 (n/a)51.49 (n/a)

rope_1_cols_2_channels_4096_tile_4096_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.19 (-21.66%)0.17 (-8.53%)0.18 (+2.09%)0.14 (-6.71%)0.02 (-38.07%)174.10 (+7.20%)150.06 (+8.11%)139.60 (-2.04%)129.80 (+27.63%)20.36 (-10.65%)
d48746f — 2026-01-12 22:17:100.24 (n/a)0.18 (n/a)0.17 (n/a)0.15 (n/a)0.04 (n/a)162.40 (n/a)138.80 (n/a)142.50 (n/a)101.70 (n/a)22.79 (n/a)

rope_1c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.65 (-39.13%)0.56 (-36.80%)0.55 (-39.97%)0.49 (-31.19%)0.06 (-56.06%)201.00 (+45.34%)177.48 (+56.31%)180.00 (+66.51%)152.10 (+64.25%)19.92 (+3.86%)
897d04e — 2026-03-06 22:12:251.06 (n/a)0.89 (n/a)0.91 (n/a)0.71 (n/a)0.15 (n/a)138.30 (n/a)113.54 (n/a)108.10 (n/a)92.60 (n/a)19.18 (n/a)

rope_1c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.40 (-35.40%)0.35 (-30.19%)0.36 (-26.58%)0.29 (-34.18%)0.04 (-42.62%)251.80 (+51.96%)212.16 (+42.95%)207.40 (+36.18%)185.30 (+54.80%)24.31 (+42.14%)
897d04e — 2026-03-06 22:12:250.62 (n/a)0.50 (n/a)0.48 (n/a)0.44 (n/a)0.07 (n/a)165.70 (n/a)148.42 (n/a)152.30 (n/a)119.70 (n/a)17.10 (n/a)

rope_2_cols_2_channels_4096_tile_2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.21 (-0.56%)0.17 (-10.62%)0.16 (-13.58%)0.13 (-13.33%)0.03 (+17.05%)183.70 (+15.39%)151.76 (+12.77%)155.20 (+15.73%)117.30 (+0.60%)24.24 (+35.73%)
d48746f — 2026-01-12 22:17:100.21 (n/a)0.19 (n/a)0.18 (n/a)0.15 (n/a)0.02 (n/a)159.20 (n/a)134.58 (n/a)134.10 (n/a)116.60 (n/a)17.86 (n/a)

rope_2c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.60 (-44.05%)0.52 (-41.56%)0.51 (-45.14%)0.45 (-20.94%)0.07 (-66.66%)218.20 (+26.49%)190.26 (+64.53%)193.10 (+82.34%)162.60 (+78.68%)24.16 (-26.99%)
897d04e — 2026-03-06 22:12:251.08 (n/a)0.90 (n/a)0.93 (n/a)0.57 (n/a)0.20 (n/a)172.50 (n/a)115.64 (n/a)105.90 (n/a)91.00 (n/a)33.09 (n/a)

rope_2c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.41 (-53.09%)0.38 (-40.18%)0.38 (-26.87%)0.33 (-32.11%)0.03 (-83.19%)223.20 (+47.33%)196.86 (+58.68%)196.50 (+36.74%)179.50 (+113.18%)16.33 (-47.25%)
897d04e — 2026-03-06 22:12:250.88 (n/a)0.63 (n/a)0.51 (n/a)0.49 (n/a)0.18 (n/a)151.50 (n/a)124.06 (n/a)143.70 (n/a)84.20 (n/a)30.96 (n/a)

rope_4_cols_2_channels_4096_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.22 (+18.92%)0.17 (+6.24%)0.17 (+5.57%)0.14 (-0.50%)0.03 (+90.77%)171.10 (+0.47%)147.40 (-4.39%)148.10 (-5.31%)110.10 (-15.89%)23.59 (+60.67%)
d48746f — 2026-01-12 22:17:100.19 (n/a)0.16 (n/a)0.16 (n/a)0.14 (n/a)0.02 (n/a)170.30 (n/a)154.16 (n/a)156.40 (n/a)130.90 (n/a)14.69 (n/a)

rope_8_cols_2_channels_4096_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.17 (-1.30%)0.15 (+6.63%)0.15 (-8.78%)0.14 (+39.68%)0.01 (-67.35%)174.90 (-28.41%)163.42 (-10.36%)167.40 (+9.63%)146.50 (+1.31%)11.19 (-76.17%)
d48746f — 2026-01-12 22:17:100.17 (n/a)0.14 (n/a)0.16 (n/a)0.10 (n/a)0.03 (n/a)244.30 (n/a)182.30 (n/a)152.70 (n/a)144.60 (n/a)46.98 (n/a)

rope_8c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.66 (-38.19%)0.51 (-42.84%)0.52 (-49.73%)0.36 (-33.78%)0.11 (-50.20%)276.10 (+51.04%)203.24 (+70.53%)190.70 (+98.85%)149.30 (+61.76%)48.65 (+25.42%)
897d04e — 2026-03-06 22:12:251.07 (n/a)0.88 (n/a)1.03 (n/a)0.54 (n/a)0.23 (n/a)182.80 (n/a)119.18 (n/a)95.90 (n/a)92.30 (n/a)38.79 (n/a)

rope_8c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.38 (-32.49%)0.32 (-33.20%)0.33 (-36.73%)0.22 (-40.63%)0.06 (-28.37%)341.90 (+68.42%)239.26 (+51.14%)222.00 (+58.12%)194.50 (+48.13%)58.84 (+87.52%)
897d04e — 2026-03-06 22:12:250.56 (n/a)0.48 (n/a)0.52 (n/a)0.36 (n/a)0.09 (n/a)203.00 (n/a)158.30 (n/a)140.40 (n/a)131.30 (n/a)31.38 (n/a)

sigmoid_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.07 (-19.28%)0.06 (-26.11%)0.06 (-36.31%)0.04 (-18.63%)0.01 (-31.27%)186.00 (+22.93%)152.74 (+33.96%)147.80 (+57.07%)115.90 (+23.96%)31.17 (+11.15%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)151.30 (n/a)114.02 (n/a)94.10 (n/a)93.50 (n/a)28.04 (n/a)

sigmoid_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.07 (-35.25%)0.05 (-42.98%)0.04 (-58.63%)0.04 (-8.48%)0.01 (-60.34%)192.90 (+9.29%)172.44 (+61.46%)186.00 (+141.87%)118.00 (+54.45%)31.56 (-29.56%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.11 (n/a)0.05 (n/a)0.03 (n/a)176.50 (n/a)106.80 (n/a)76.90 (n/a)76.40 (n/a)44.81 (n/a)

sigmoid_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.07 (-20.09%)0.05 (-32.61%)0.04 (-46.73%)0.03 (-28.38%)0.02 (+5.86%)239.60 (+39.63%)174.28 (+54.34%)188.10 (+87.72%)113.00 (+25.14%)56.20 (+68.82%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.08 (n/a)0.05 (n/a)0.02 (n/a)171.60 (n/a)112.92 (n/a)100.20 (n/a)90.30 (n/a)33.29 (n/a)

sigmoid_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.05 (-54.57%)0.04 (-52.77%)0.04 (-57.89%)0.03 (-13.39%)0.00 (-83.25%)246.00 (+15.44%)202.88 (+85.82%)198.10 (+137.53%)181.90 (+120.22%)25.11 (-56.79%)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.09 (n/a)0.10 (n/a)0.04 (n/a)0.03 (n/a)213.10 (n/a)109.18 (n/a)83.40 (n/a)82.60 (n/a)58.09 (n/a)

silu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.07 (-32.98%)0.05 (-38.15%)0.05 (-52.40%)0.03 (-24.00%)0.02 (-49.70%)248.20 (+31.60%)170.74 (+51.37%)165.20 (+110.18%)113.00 (+49.27%)52.91 (+3.67%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.08 (n/a)0.10 (n/a)0.04 (n/a)0.03 (n/a)188.60 (n/a)112.80 (n/a)78.60 (n/a)75.70 (n/a)51.04 (n/a)

silu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.07 (-30.29%)0.05 (-34.58%)0.05 (-39.39%)0.04 (-21.66%)0.01 (-43.08%)184.30 (+27.63%)168.68 (+50.63%)179.50 (+64.98%)124.90 (+43.40%)25.06 (+4.12%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.02 (n/a)144.40 (n/a)111.98 (n/a)108.80 (n/a)87.10 (n/a)24.06 (n/a)

silu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.05 (-41.61%)0.04 (-44.89%)0.04 (-47.02%)0.04 (-28.99%)0.01 (-54.88%)233.70 (+40.87%)195.86 (+76.99%)186.90 (+88.79%)153.80 (+71.27%)32.74 (+5.11%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.08 (n/a)0.05 (n/a)0.02 (n/a)165.90 (n/a)110.66 (n/a)99.00 (n/a)89.80 (n/a)31.15 (n/a)

silu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.06 (-38.55%)0.04 (-48.23%)0.04 (-58.52%)0.03 (-23.62%)0.01 (-52.63%)239.40 (+30.96%)192.66 (+83.63%)210.30 (+141.17%)132.60 (+62.70%)41.64 (-4.57%)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.09 (n/a)0.09 (n/a)0.04 (n/a)0.02 (n/a)182.80 (n/a)104.92 (n/a)87.20 (n/a)81.50 (n/a)43.64 (n/a)

softmax_1_cols_2_channels_4096_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.19 (-8.94%)0.17 (-8.66%)0.17 (-19.83%)0.14 (+4.32%)0.02 (-35.13%)115.80 (-4.14%)99.26 (+7.47%)98.60 (+24.65%)85.00 (+9.82%)13.65 (-30.86%)
84d3478 — 2026-02-17 23:25:130.21 (n/a)0.18 (n/a)0.21 (n/a)0.14 (n/a)0.04 (n/a)120.80 (n/a)92.36 (n/a)79.10 (n/a)77.40 (n/a)19.75 (n/a)

softmax_2_cols_2_channels_32768_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.90 (n/a)0.71 (n/a)0.74 (n/a)0.57 (n/a)0.13 (n/a)229.60 (n/a)189.12 (n/a)176.90 (n/a)145.90 (n/a)34.97 (n/a)

softmax_2_cols_2_channels_32768_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.91 (n/a)0.68 (n/a)0.63 (n/a)0.55 (n/a)0.14 (n/a)236.30 (n/a)199.98 (n/a)208.10 (n/a)143.50 (n/a)36.24 (n/a)

softmax_2_cols_2_channels_4096_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.20 (-0.87%)0.16 (+6.57%)0.17 (-9.55%)0.11 (+29.81%)0.03 (-44.34%)145.70 (-22.99%)104.68 (-15.00%)95.00 (+10.59%)83.80 (+0.84%)24.16 (-54.23%)
84d3478 — 2026-02-17 23:25:130.20 (n/a)0.15 (n/a)0.19 (n/a)0.09 (n/a)0.06 (n/a)189.20 (n/a)123.16 (n/a)85.90 (n/a)83.10 (n/a)52.80 (n/a)

softmax_2_cols_2_channels_4096_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.25 (+23.53%)0.18 (+9.42%)0.18 (-3.86%)0.11 (+2.99%)0.05 (+19.24%)144.10 (-2.90%)98.98 (-8.18%)92.10 (+4.07%)64.30 (-19.02%)29.58 (-7.13%)
84d3478 — 2026-02-17 23:25:130.21 (n/a)0.16 (n/a)0.19 (n/a)0.11 (n/a)0.04 (n/a)148.40 (n/a)107.80 (n/a)88.50 (n/a)79.40 (n/a)31.85 (n/a)

softmax_4_cols_4_channels_32768_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.71 (n/a)0.48 (n/a)0.54 (n/a)0.30 (n/a)0.18 (n/a)432.70 (n/a)305.66 (n/a)240.60 (n/a)184.80 (n/a)117.50 (n/a)

swiglu

No metrics available.

swiglu_decode_1x2048x2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:49n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)5187.20 (n/a)4720.84 (n/a)4750.90 (n/a)4015.10 (n/a)390.23 (n/a)

swiglu_decode_1x2048x2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.00 (-0.21%)0.00 (+7.74%)0.00 (+0.37%)0.00 (+8.84%)0.00 (-19.58%)5170.80 (-8.12%)3761.80 (-8.70%)3408.60 (-0.36%)3407.30 (+0.21%)787.66 (-22.98%)
84d3478 — 2026-02-17 23:25:130.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)5627.70 (n/a)4120.44 (n/a)3421.00 (n/a)3400.30 (n/a)1022.61 (n/a)

swiglu_prefill_256x2048x2048

No metrics available.

tanh_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.07 (-23.98%)0.05 (-32.96%)0.05 (-35.01%)0.04 (-35.30%)0.01 (+10.45%)200.50 (+54.59%)161.78 (+51.88%)161.60 (+53.90%)125.40 (+31.58%)31.19 (+123.63%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)129.70 (n/a)106.52 (n/a)105.00 (n/a)95.30 (n/a)13.95 (n/a)

tanh_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.05 (-50.67%)0.04 (-53.66%)0.04 (-58.53%)0.03 (-44.10%)0.01 (-56.70%)286.10 (+78.92%)199.08 (+109.73%)183.60 (+141.26%)151.40 (+102.68%)55.76 (+51.83%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.11 (n/a)0.05 (n/a)0.02 (n/a)159.90 (n/a)94.92 (n/a)76.10 (n/a)74.70 (n/a)36.73 (n/a)

tanh_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.07 (-18.93%)0.05 (-32.72%)0.05 (-35.19%)0.04 (-41.77%)0.01 (+37.46%)221.50 (+71.71%)172.46 (+53.16%)164.50 (+54.32%)124.90 (+23.30%)38.23 (+194.77%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.07 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)129.00 (n/a)112.60 (n/a)106.60 (n/a)101.30 (n/a)12.97 (n/a)

tanh_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.05 (-35.90%)0.04 (-40.98%)0.05 (-43.59%)0.03 (-48.56%)0.01 (-10.52%)259.90 (+94.39%)189.90 (+72.64%)180.70 (+77.33%)155.00 (+56.09%)40.86 (+182.01%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)133.70 (n/a)110.00 (n/a)101.90 (n/a)99.30 (n/a)14.49 (n/a)

transpose_2048_M_64_N_1_cols_1_channels_64_m_64_n_8_s

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:493.91 (-0.52%)2.99 (-18.85%)3.08 (-15.78%)1.99 (-42.43%)0.69 (+353.72%)263.70 (+73.72%)184.24 (+29.11%)170.20 (+18.69%)134.20 (+0.52%)48.37 (+720.25%)
130b6ea — 2025-12-05 21:15:493.93 (n/a)3.68 (n/a)3.66 (n/a)3.45 (n/a)0.15 (n/a)151.80 (n/a)142.70 (n/a)143.40 (n/a)133.50 (n/a)5.90 (n/a)

transpose_2048_M_64_N_1_cols_1_channels_64_m_64_n_8_s0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:254.43 (+0.85%)3.73 (+3.83%)4.02 (-1.30%)2.97 (+16.77%)0.65 (-24.99%)176.70 (-14.35%)144.46 (-6.24%)130.40 (+1.32%)118.40 (-0.84%)26.60 (-35.16%)
84d3478 — 2026-02-17 23:25:134.39 (n/a)3.59 (n/a)4.07 (n/a)2.54 (n/a)0.87 (n/a)206.30 (n/a)154.08 (n/a)128.70 (n/a)119.40 (n/a)41.02 (n/a)

transpose_2048_M_64_N_1_cols_2_channels_64_m_64_n_8_s

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:494.19 (-18.01%)3.34 (-13.30%)3.28 (-3.64%)2.39 (-27.16%)0.66 (-7.65%)219.10 (+37.28%)162.32 (+15.79%)160.00 (+3.76%)125.20 (+22.03%)35.16 (+54.89%)
130b6ea — 2025-12-05 21:15:495.11 (n/a)3.85 (n/a)3.40 (n/a)3.29 (n/a)0.71 (n/a)159.60 (n/a)140.18 (n/a)154.20 (n/a)102.60 (n/a)22.70 (n/a)

transpose_2048_M_64_N_1_cols_2_channels_64_m_64_n_8_s0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:255.27 (-14.18%)4.35 (-2.49%)4.45 (+2.15%)2.78 (-22.18%)0.98 (-3.38%)188.60 (+28.47%)126.98 (+4.20%)117.90 (-2.08%)99.40 (+16.53%)36.02 (+50.15%)
84d3478 — 2026-02-17 23:25:136.14 (n/a)4.46 (n/a)4.35 (n/a)3.57 (n/a)1.02 (n/a)146.80 (n/a)121.86 (n/a)120.40 (n/a)85.30 (n/a)23.99 (n/a)

weighted_rms_norm_1_cols_2_channels_2048_weights_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.09 (+15.48%)0.07 (-3.52%)0.06 (-16.12%)0.06 (+22.59%)0.02 (+10.47%)207.20 (+22.39%)187.34 (+54.60%)201.90 (+78.83%)129.70 (+29.83%)32.49 (+13.96%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.07 (n/a)0.07 (n/a)0.05 (n/a)0.01 (n/a)169.30 (n/a)121.18 (n/a)112.90 (n/a)99.90 (n/a)28.51 (n/a)

weighted_rms_norm_2_cols_2_channels_2048_weights_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.09 (-19.09%)0.07 (-28.61%)0.07 (-31.52%)0.05 (-20.46%)0.02 (-19.80%)220.30 (+57.13%)158.74 (+74.82%)143.80 (+82.49%)120.40 (+54.56%)42.32 (+53.23%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.10 (n/a)0.10 (n/a)0.06 (n/a)0.02 (n/a)140.20 (n/a)90.80 (n/a)78.80 (n/a)77.90 (n/a)27.62 (n/a)

weighted_rms_norm_4_cols_2_channels_2048_weights_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.09 (-19.21%)0.06 (-39.95%)0.05 (-48.21%)0.04 (-30.49%)0.02 (-4.82%)205.00 (+61.80%)169.94 (+90.56%)173.10 (+117.19%)107.50 (+39.25%)37.71 (+78.76%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.10 (n/a)0.10 (n/a)0.06 (n/a)0.02 (n/a)126.70 (n/a)89.18 (n/a)79.70 (n/a)77.20 (n/a)21.10 (n/a)

weighted_rms_norm_8_cols_2_channels_2048_weights_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
2f954d5 — 2026-03-18 18:02:490.05 (-33.56%)0.04 (-49.43%)0.04 (-51.84%)0.02 (-49.21%)0.01 (-17.57%)371.90 (+109.17%)262.02 (+119.41%)229.40 (+120.58%)164.50 (+59.86%)86.66 (+164.84%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.07 (n/a)0.08 (n/a)0.05 (n/a)0.01 (n/a)177.80 (n/a)119.42 (n/a)104.00 (n/a)102.90 (n/a)32.72 (n/a)

@github-actions
Copy link
Contributor

📊 Test Results for Test Example Applications

2f954d5 (2026_03_18_18_15_20)

IRONCLAD

Tested on 2026_03_18_18_15_20 at commit 2f954d5.

Test Checks TTFT (mean)TPS (mean)
llama_3.2_1b_prompt_1024_tokens_40 ✅ 1/1 2.37 n/a
📈 Trends (vs main branch) for Test Example Applications

2f954d5 (2026_03_18_18_15_20)

IRONCLAD Trends

llama_3.2_1b

Commit/Date Num Tokens (max)Num Tokens (mean)Num Tokens (median)Num Tokens (min)Num Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)Total (max)Total (mean)Total (median)Total (min)Total (stddev)
130b6ea — 2025-12-05 21:33:1240.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.71 (-0.42%)4.64 (-0.09%)4.64 (+0.65%)4.55 (-0.22%)0.05 (-17.66%)4.41 (-0.34%)4.39 (-0.19%)4.38 (-0.33%)4.37 (-0.15%)0.01 (-25.90%)12.96 (-0.00%)12.80 (+0.07%)12.80 (-0.23%)12.67 (+0.44%)0.09 (-21.12%)
0a6c11c — 2025-12-03 23:35:1540.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.73 (n/a)4.64 (n/a)4.61 (n/a)4.56 (n/a)0.06 (n/a)4.42 (n/a)4.40 (n/a)4.40 (n/a)4.37 (n/a)0.02 (n/a)12.96 (n/a)12.79 (n/a)12.83 (n/a)12.62 (n/a)0.12 (n/a)

llama_3.2_1b_prompt_1024_tokens_40

Commit/Date TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
2f954d5 — 2026-03-18 18:13:182.37 (n/a)2.37 (n/a)2.37 (n/a)2.37 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_13_tokens_1

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:071.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.62 (-4.45%)0.62 (-4.45%)0.62 (-4.45%)0.62 (-4.45%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:231.00 (n/a)1.00 (n/a)1.00 (n/a)1.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.65 (n/a)0.65 (n/a)0.65 (n/a)0.65 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_13_tokens_40

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:0740.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.30 (+2.38%)4.30 (+2.38%)4.30 (+2.38%)4.30 (+2.38%)0.00 (n/a)0.61 (-2.49%)0.61 (-2.49%)0.61 (-2.49%)0.61 (-2.49%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:2340.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.20 (n/a)4.20 (n/a)4.20 (n/a)4.20 (n/a)0.00 (n/a)0.63 (n/a)0.63 (n/a)0.63 (n/a)0.63 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_2048_tokens_1

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:071.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:231.00 (n/a)1.00 (n/a)1.00 (n/a)1.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_2048_tokens_40

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:0740.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)0.00 (n/a)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:2340.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)0.00 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)0.00 (n/a)

@andrej
Copy link
Collaborator Author

andrej commented Mar 18, 2026

Required fix for runlist-based tests: MLIR-AIE/PR 2991

@andrej andrej force-pushed the alt-simplifying-refactor branch from a4e9d52 to c509661 Compare March 20, 2026 17:07
@github-actions
Copy link
Contributor

📊 Test Results for Small Benchmark/Test Suite

27cd2f3 (2026_03_20_17_16_03)

IRONCLAD

Tested on 2026_03_20_17_16_03 at commit 27cd2f3.

Test Checks Latency (mean)Bandwidth (mean)GFLOP/s (mean)
axpy_1_cols_2_channels_2048_tile_2048_3.0 ✅ 5/5 145.24 0.09 n/a
axpy_2_cols_2_channels_2048_tile_1024_3.0 ✅ 5/5 189.94 0.07 n/a
axpy_4_cols_2_channels_2048_tile_512_3.0 ✅ 5/5 181.86 0.07 n/a
axpy_8_cols_2_channels_2048_tile_256_3.0 ✅ 5/5 207.04 0.06 n/a
dequant_1_cols_1_channels_2048_tile_2048 ✅ 5/5 190.90 0.03 n/a
dequant_1_cols_2_channels_2048_tile_1024 ✅ 5/5 177.20 0.03 n/a
dequant_2_cols_1_channels_2048_tile_1024 ✅ 5/5 185.50 0.03 n/a
dequant_2_cols_2_channels_2048_tile_512 ✅ 5/5 192.36 0.03 n/a
dequant_4_cols_1_channels_2048_tile_512 ✅ 5/5 204.34 0.03 n/a
dequant_4_cols_2_channels_2048_tile_256 ✅ 5/5 193.08 0.03 n/a
dequant_8_cols_1_channels_2048_tile_256 ✅ 5/5 184.84 0.03 n/a
dequant_8_cols_2_channels_2048_tile_128 ✅ 5/5 235.84 0.02 n/a
eltwise_add_1_cols_2_channels_2048_tile_2048 ✅ 5/5 194.08 0.07 n/a
eltwise_add_2_cols_2_channels_2048_tile_1024 ✅ 5/5 168.48 0.07 n/a
eltwise_add_4_cols_2_channels_2048_tile_512 ✅ 5/5 202.18 0.06 n/a
eltwise_add_8_cols_2_channels_2048_tile_256 ✅ 5/5 263.26 0.05 n/a
eltwise_mul_1_cols_2_channels_2048_tile_2048 ✅ 5/5 214.56 0.06 n/a
eltwise_mul_2_cols_2_channels_2048_tile_1024 ✅ 5/5 191.78 0.07 n/a
eltwise_mul_4_cols_2_channels_2048_tile_512 ✅ 5/5 221.98 0.06 n/a
eltwise_mul_8_cols_2_channels_2048_tile_256 ✅ 5/5 242.08 0.05 n/a
gelu_1_cols_1_channels_2048_tile_2048 ✅ 5/5 171.34 0.05 n/a
gelu_1_cols_2_channels_2048_tile_1024 ✅ 5/5 222.70 0.04 n/a
gelu_2_cols_1_channels_2048_tile_1024 ✅ 5/5 204.68 0.04 n/a
gelu_2_cols_2_channels_2048_tile_512 ✅ 5/5 186.34 0.05 n/a
gelu_4_cols_1_channels_2048_tile_512 ✅ 5/5 215.62 0.04 n/a
gelu_4_cols_2_channels_2048_tile_256 ✅ 5/5 183.50 0.05 n/a
gelu_8_cols_1_channels_2048_tile_256 ✅ 5/5 179.20 0.05 n/a
gelu_8_cols_2_channels_2048_tile_128 ✅ 5/5 225.40 0.04 n/a
gemm_1792x896x1152_64x32x48_8cols_ccolmaj ✅ 5/5 2133.76 4.44 n/a
gemm_192x384x64_48x96x16_4cols ✅ 5/5 320.84 0.73 n/a
gemm_192x384x64_48x96x16_4cols_bcolmaj_ccolmaj ✅ 5/5 214.38 1.10 n/a
gemm_2048x2048x2048_64x64x64_1cols ✅ 5/5 49284.16 0.51 n/a
gemm_2048x2048x2048_64x64x64_2cols_bcolmaj ✅ 5/5 28669.54 0.88 n/a
gemm_2048x2048x2048_64x64x64_8cols_bcolmaj_ccolmaj ✅ 5/5 7483.34 3.36 n/a
gemm_384x1536x1792_32x48x64_4cols_bcolmaj ✅ 5/5 2295.02 3.54 n/a
gemm_896x1792x640_32x64x80_8cols_ccolmaj ✅ 5/5 1598.22 4.31 n/a
layer_norm_1_cols_1_channels_2048_tile_2048 ✅ 5/5 178.76 0.05 n/a
layer_norm_1_cols_2_channels_2048_tile_1024 ✅ 5/5 180.34 0.05 n/a
layer_norm_2_cols_1_channels_2048_tile_1024 ✅ 5/5 169.04 0.05 n/a
layer_norm_2_cols_2_channels_2048_tile_512 ✅ 5/5 194.44 0.05 n/a
layer_norm_4_cols_1_channels_2048_tile_512 ✅ 5/5 179.96 0.05 n/a
layer_norm_4_cols_2_channels_2048_tile_256 ✅ 5/5 226.88 0.04 n/a
layer_norm_8_cols_1_channels_2048_tile_256 ✅ 5/5 187.36 0.05 n/a
layer_norm_8_cols_2_channels_2048_tile_128 ✅ 5/5 216.54 0.04 n/a
matrix_vector_mul_128x128_32tsi_128tso_1col ✅ 5/5 n/a 0.18 n/a
matrix_vector_mul_2048x8192_1tsi_1024tso_2col ✅ 5/5 n/a 23.27 n/a
matrix_vector_mul_2048x8192_1tsi_2048tso_1col ✅ 5/5 n/a 12.98 n/a
matrix_vector_mul_2048x8192_1tsi_256tso_8col ✅ 5/5 n/a 40.49 n/a
matrix_vector_mul_2048x8192_1tsi_512tso_4col ✅ 5/5 n/a 38.51 n/a
matrix_vector_mul_8192x2048_4tsi_1024tso_1col ✅ 5/5 n/a 13.04 n/a
matrix_vector_mul_8192x2048_4tsi_1024tso_2col ✅ 5/5 n/a 23.97 n/a
matrix_vector_mul_8192x2048_4tsi_1024tso_4col ✅ 5/5 n/a 37.22 n/a
matrix_vector_mul_8192x2048_4tsi_1024tso_8col ✅ 5/5 n/a 41.60 n/a
mem_copy_16_cores_2_chans_2048_tile_128_False ✅ 5/5 252.10 0.04 n/a
mem_copy_1_cores_1_chans_2048_tile_2048_False ✅ 5/5 156.62 0.05 n/a
mem_copy_2_cores_1_chans_2048_tile_1024_False ✅ 5/5 165.42 0.05 n/a
mem_copy_2_cores_2_chans_2048_tile_1024_False ✅ 5/5 212.06 0.04 n/a
mem_copy_4_cores_1_chans_2048_tile_512_False ✅ 5/5 179.50 0.05 n/a
mem_copy_4_cores_2_chans_2048_tile_512_False ✅ 5/5 179.54 0.05 n/a
mem_copy_8_cores_1_chans_2048_tile_256_False ✅ 5/5 200.18 0.04 n/a
mem_copy_8_cores_2_chans_2048_tile_256_False ✅ 5/5 213.08 0.04 n/a
mha ✅ 5/5 47504.24 0.18 n/a
relu_1_cols_1_channels_2048_tile_2048 ✅ 5/5 171.00 0.05 n/a
relu_2_cols_1_channels_2048_tile_1024 ✅ 5/5 165.38 0.05 n/a
relu_4_cols_1_channels_2048_tile_512 ✅ 5/5 193.56 0.04 n/a
relu_8_cols_1_channels_2048_tile_256 ✅ 5/5 174.38 0.05 n/a
rms_norm_1_cols_1_channels_2048_tile_2048 ✅ 5/5 170.52 0.05 n/a
rms_norm_1_cols_2_channels_2048_tile_1024 ✅ 5/5 169.38 0.05 n/a
rms_norm_2_cols_1_channels_2048_tile_1024 ✅ 5/5 160.88 0.05 n/a
rms_norm_2_cols_2_channels_2048_tile_512 ✅ 5/5 186.06 0.05 n/a
rms_norm_4_cols_1_channels_2048_tile_512 ✅ 5/5 168.60 0.05 n/a
rms_norm_4_cols_2_channels_2048_tile_256 ✅ 5/5 199.26 0.04 n/a
rms_norm_8_cols_1_channels_2048_tile_256 ✅ 5/5 195.54 0.04 n/a
rms_norm_8_cols_2_channels_2048_tile_128 ✅ 5/5 227.02 0.04 n/a
rope_1c_32rows_512cols_32arows_0m ✅ 5/5 157.78 0.67 n/a
rope_1c_32rows_512cols_8arows_0m ✅ 5/5 165.28 0.47 n/a
rope_2c_32rows_512cols_32arows_0m ✅ 5/5 160.98 0.64 n/a
rope_2c_32rows_512cols_8arows_0m ✅ 5/5 159.74 0.47 n/a
rope_8c_32rows_512cols_32arows_0m ✅ 5/5 206.04 0.53 n/a
rope_8c_32rows_512cols_8arows_0m ✅ 5/5 194.68 0.40 n/a
sigmoid_1_cols_1_channels_2048_tile_2048 ✅ 5/5 174.10 0.05 n/a
sigmoid_2_cols_1_channels_2048_tile_1024 ✅ 5/5 199.38 0.04 n/a
sigmoid_4_cols_1_channels_2048_tile_512 ✅ 5/5 180.26 0.05 n/a
sigmoid_8_cols_1_channels_2048_tile_256 ✅ 5/5 197.88 0.04 n/a
silu_1_cols_1_channels_2048_tile_2048 ✅ 5/5 163.82 0.05 n/a
silu_2_cols_1_channels_2048_tile_1024 ✅ 5/5 155.16 0.05 n/a
silu_4_cols_1_channels_2048_tile_512 ✅ 5/5 200.28 0.04 n/a
silu_8_cols_1_channels_2048_tile_256 ✅ 5/5 172.24 0.05 n/a
softmax_2_cols_2_channels_32768_tile_1024 ✅ 5/5 191.46 0.73 n/a
softmax_2_cols_2_channels_32768_tile_512 ✅ 5/5 186.26 0.77 n/a
softmax_4_cols_4_channels_32768_tile_2048 ✅ 5/5 224.02 0.60 n/a
swiglu_decode_1x2048x2048 ✅ 5/5 n/a n/a n/a
swiglu_prefill_256x2048x2048 ✅ 5/5 n/a n/a n/a
tanh_1_cols_1_channels_2048_tile_2048 ✅ 5/5 167.64 0.05 n/a
tanh_2_cols_1_channels_2048_tile_1024 ✅ 5/5 190.58 0.04 n/a
tanh_4_cols_1_channels_2048_tile_512 ✅ 5/5 165.44 0.05 n/a
tanh_8_cols_1_channels_2048_tile_256 ✅ 5/5 197.56 0.04 n/a
transpose_2048_M_64_N_1_cols_1_channels_64_m_64_n_8_s ✅ 5/5 229.08 2.36 n/a
transpose_2048_M_64_N_1_cols_2_channels_64_m_64_n_8_s ✅ 5/5 203.28 2.61 n/a
weighted_rms_norm_1_cols_2_channels_2048_weights_2048 ✅ 5/5 191.50 0.07 n/a
weighted_rms_norm_2_cols_2_channels_2048_weights_1024 ✅ 5/5 191.98 0.05 n/a
weighted_rms_norm_4_cols_2_channels_2048_weights_512 ✅ 5/5 254.86 0.04 n/a
weighted_rms_norm_8_cols_2_channels_2048_weights_256 ✅ 5/5 214.32 0.04 n/a
📈 Trends (vs main branch) for Small Benchmark/Test Suite

27cd2f3 (2026_03_20_17_16_03)

IRONCLAD Trends

axpy_1_cols_2_channels_2048_tile_2048_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.13 (+2.27%)0.09 (-14.06%)0.08 (-23.06%)0.06 (-16.61%)0.03 (+49.17%)194.80 (+19.95%)145.24 (+20.07%)152.10 (+30.00%)92.30 (-2.22%)38.74 (+67.62%)
130b6ea — 2025-12-05 21:15:490.13 (n/a)0.11 (n/a)0.10 (n/a)0.08 (n/a)0.02 (n/a)162.40 (n/a)120.96 (n/a)117.00 (n/a)94.40 (n/a)23.11 (n/a)

axpy_1_cols_2_channels_2048_tile_2048_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.16 (-3.31%)0.14 (-1.29%)0.15 (+16.18%)0.10 (-6.80%)0.03 (+3.89%)125.70 (+7.25%)93.56 (+2.12%)83.50 (-13.92%)75.00 (+3.45%)22.38 (+18.09%)
84d3478 — 2026-02-17 23:25:130.17 (n/a)0.14 (n/a)0.13 (n/a)0.10 (n/a)0.03 (n/a)117.20 (n/a)91.62 (n/a)97.00 (n/a)72.50 (n/a)18.95 (n/a)

axpy_2_cols_2_channels_2048_tile_1024_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.09 (-18.25%)0.07 (-26.06%)0.06 (-25.97%)0.05 (-42.61%)0.02 (+52.02%)264.90 (+74.28%)189.94 (+40.22%)192.90 (+35.08%)140.70 (+22.35%)48.63 (+223.45%)
130b6ea — 2025-12-05 21:15:490.11 (n/a)0.09 (n/a)0.09 (n/a)0.08 (n/a)0.01 (n/a)152.00 (n/a)135.46 (n/a)142.80 (n/a)115.00 (n/a)15.03 (n/a)

axpy_2_cols_2_channels_2048_tile_1024_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (-0.73%)0.11 (-7.43%)0.11 (-8.63%)0.07 (-27.71%)0.02 (+50.39%)182.70 (+38.30%)120.86 (+11.97%)110.80 (+9.38%)96.10 (+0.73%)35.16 (+122.88%)
84d3478 — 2026-02-17 23:25:130.13 (n/a)0.12 (n/a)0.12 (n/a)0.09 (n/a)0.02 (n/a)132.10 (n/a)107.94 (n/a)101.30 (n/a)95.40 (n/a)15.78 (n/a)

axpy_4_cols_2_channels_2048_tile_512_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.08 (-29.00%)0.07 (-30.11%)0.06 (-36.24%)0.06 (-20.75%)0.01 (-37.09%)210.30 (+26.15%)181.86 (+41.22%)191.00 (+56.81%)146.40 (+40.90%)25.48 (+11.02%)
130b6ea — 2025-12-05 21:15:490.12 (n/a)0.10 (n/a)0.10 (n/a)0.07 (n/a)0.02 (n/a)166.70 (n/a)128.78 (n/a)121.80 (n/a)103.90 (n/a)22.95 (n/a)

axpy_4_cols_2_channels_2048_tile_512_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (-10.91%)0.11 (-7.68%)0.12 (-0.73%)0.06 (-17.94%)0.03 (+0.99%)194.10 (+21.85%)124.52 (+10.29%)103.30 (+0.68%)97.20 (+12.24%)40.32 (+39.15%)
84d3478 — 2026-02-17 23:25:130.14 (n/a)0.11 (n/a)0.12 (n/a)0.08 (n/a)0.03 (n/a)159.30 (n/a)112.90 (n/a)102.60 (n/a)86.60 (n/a)28.97 (n/a)

axpy_8_cols_2_channels_2048_tile_256_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.08 (-21.38%)0.06 (-25.20%)0.06 (-33.89%)0.05 (-27.18%)0.01 (+8.74%)253.20 (+37.31%)207.04 (+35.16%)217.90 (+51.21%)161.60 (+27.14%)39.14 (+84.08%)
130b6ea — 2025-12-05 21:15:490.10 (n/a)0.08 (n/a)0.09 (n/a)0.07 (n/a)0.01 (n/a)184.40 (n/a)153.18 (n/a)144.10 (n/a)127.10 (n/a)21.26 (n/a)

axpy_8_cols_2_channels_2048_tile_256_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (+0.01%)0.11 (+6.10%)0.11 (+0.22%)0.08 (-0.33%)0.02 (-9.83%)156.30 (+0.32%)115.26 (-6.40%)112.70 (-0.18%)95.50 (+0.00%)24.34 (-10.22%)
84d3478 — 2026-02-17 23:25:130.13 (n/a)0.10 (n/a)0.11 (n/a)0.08 (n/a)0.02 (n/a)155.80 (n/a)123.14 (n/a)112.90 (n/a)95.50 (n/a)27.11 (n/a)

dequant_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.03 (-24.64%)0.03 (-24.62%)0.03 (-27.58%)0.02 (-34.16%)0.01 (+21.00%)243.70 (+51.93%)190.90 (+35.27%)188.40 (+38.12%)151.60 (+32.63%)40.27 (+131.82%)
130b6ea — 2025-12-05 21:15:490.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.00 (n/a)160.40 (n/a)141.12 (n/a)136.40 (n/a)114.30 (n/a)17.37 (n/a)

dequant_1_cols_1_channels_2048_tile_2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-13.00%)0.04 (-0.87%)0.04 (+1.16%)0.03 (-3.17%)0.01 (-31.33%)158.40 (+3.26%)128.12 (-0.19%)123.80 (-1.12%)110.40 (+14.88%)18.36 (-15.65%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)153.40 (n/a)128.36 (n/a)125.20 (n/a)96.10 (n/a)21.77 (n/a)

dequant_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.04 (-26.71%)0.03 (-39.29%)0.03 (-49.23%)0.03 (-35.30%)0.01 (-4.17%)206.50 (+54.57%)177.20 (+67.04%)187.40 (+97.06%)119.90 (+36.41%)35.99 (+101.07%)
130b6ea — 2025-12-05 21:15:490.06 (n/a)0.05 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)133.60 (n/a)106.08 (n/a)95.10 (n/a)87.90 (n/a)17.90 (n/a)

dequant_1_cols_2_channels_2048_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (-3.74%)0.05 (-12.63%)0.05 (-5.21%)0.03 (-33.47%)0.01 (+52.60%)162.70 (+50.37%)110.84 (+18.98%)98.80 (+5.44%)81.50 (+3.95%)31.02 (+149.23%)
84d3478 — 2026-02-17 23:25:130.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)108.20 (n/a)93.16 (n/a)93.70 (n/a)78.40 (n/a)12.45 (n/a)

dequant_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.04 (-39.16%)0.03 (-44.18%)0.03 (-40.15%)0.02 (-41.33%)0.01 (-16.94%)244.40 (+70.43%)185.50 (+83.63%)163.10 (+67.11%)129.70 (+64.39%)54.08 (+139.68%)
130b6ea — 2025-12-05 21:15:490.07 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)143.40 (n/a)101.02 (n/a)97.60 (n/a)78.90 (n/a)22.56 (n/a)

dequant_2_cols_1_channels_2048_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (+8.90%)0.05 (+1.31%)0.05 (-1.85%)0.03 (+6.35%)0.01 (+7.11%)152.30 (-5.93%)115.08 (-1.20%)100.80 (+1.92%)88.10 (-8.23%)27.20 (-4.42%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)161.90 (n/a)116.48 (n/a)98.90 (n/a)96.00 (n/a)28.46 (n/a)

dequant_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.03 (-32.44%)0.03 (-34.76%)0.02 (-48.20%)0.02 (-27.60%)0.00 (-31.09%)220.50 (+38.07%)192.36 (+52.28%)210.50 (+93.12%)158.10 (+48.03%)31.48 (+40.01%)
130b6ea — 2025-12-05 21:15:490.05 (n/a)0.04 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)159.70 (n/a)126.32 (n/a)109.00 (n/a)106.80 (n/a)22.48 (n/a)

dequant_2_cols_2_channels_2048_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.07 (+27.96%)0.05 (+1.57%)0.04 (-18.06%)0.03 (+10.23%)0.01 (+22.79%)164.70 (-9.31%)119.12 (-1.39%)120.20 (+22.03%)75.60 (-21.82%)31.61 (-13.80%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)181.60 (n/a)120.80 (n/a)98.50 (n/a)96.70 (n/a)36.67 (n/a)

dequant_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.03 (-50.36%)0.03 (-42.80%)0.03 (-45.23%)0.02 (-29.76%)0.00 (-71.63%)222.20 (+42.34%)204.34 (+69.07%)209.80 (+82.59%)177.90 (+101.47%)19.82 (-18.77%)
130b6ea — 2025-12-05 21:15:490.06 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)156.10 (n/a)120.86 (n/a)114.90 (n/a)88.30 (n/a)24.40 (n/a)

dequant_4_cols_1_channels_2048_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-2.55%)0.04 (-12.79%)0.05 (-16.25%)0.03 (-12.85%)0.01 (-1.20%)164.60 (+14.78%)122.14 (+15.14%)115.90 (+19.48%)97.50 (+2.63%)25.08 (+19.84%)
84d3478 — 2026-02-17 23:25:130.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)143.40 (n/a)106.08 (n/a)97.00 (n/a)95.00 (n/a)20.93 (n/a)

dequant_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.04 (-16.41%)0.03 (-28.23%)0.03 (-36.20%)0.02 (-27.75%)0.01 (+14.76%)232.70 (+38.43%)193.08 (+42.24%)207.20 (+56.73%)128.50 (+19.65%)43.65 (+90.58%)
130b6ea — 2025-12-05 21:15:490.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)168.10 (n/a)135.74 (n/a)132.20 (n/a)107.40 (n/a)22.91 (n/a)

dequant_4_cols_2_channels_2048_tile_256_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (+20.83%)0.05 (+8.21%)0.04 (+1.42%)0.03 (+17.39%)0.01 (+14.84%)166.40 (-14.80%)122.34 (-7.97%)127.70 (-1.39%)84.60 (-17.22%)30.47 (-19.21%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)195.30 (n/a)132.94 (n/a)129.50 (n/a)102.20 (n/a)37.71 (n/a)

dequant_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.04 (-3.50%)0.03 (-22.29%)0.03 (-32.21%)0.02 (-20.41%)0.01 (+60.80%)214.10 (+25.65%)184.84 (+31.92%)200.30 (+47.50%)124.00 (+3.68%)36.66 (+104.37%)
130b6ea — 2025-12-05 21:15:490.04 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.00 (n/a)170.40 (n/a)140.12 (n/a)135.80 (n/a)119.60 (n/a)17.94 (n/a)

dequant_8_cols_1_channels_2048_tile_256_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-25.19%)0.04 (-21.02%)0.04 (-19.70%)0.04 (-19.76%)0.00 (-37.92%)142.70 (+24.63%)126.44 (+26.24%)125.70 (+24.46%)115.50 (+33.68%)10.61 (+3.97%)
84d3478 — 2026-02-17 23:25:130.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)114.50 (n/a)100.16 (n/a)101.00 (n/a)86.40 (n/a)10.20 (n/a)

dequant_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.03 (-48.84%)0.02 (-34.67%)0.02 (-29.66%)0.02 (-24.09%)0.00 (-76.59%)261.10 (+31.74%)235.84 (+47.40%)236.30 (+42.18%)208.10 (+95.40%)18.84 (-36.91%)
130b6ea — 2025-12-05 21:15:490.05 (n/a)0.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)198.20 (n/a)160.00 (n/a)166.20 (n/a)106.50 (n/a)29.86 (n/a)

dequant_8_cols_2_channels_2048_tile_128_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.04 (-26.69%)0.04 (-11.93%)0.04 (-16.99%)0.03 (+27.28%)0.01 (-62.60%)193.10 (-21.41%)147.68 (+2.96%)136.10 (+20.44%)135.40 (+36.49%)25.41 (-58.96%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.05 (n/a)0.02 (n/a)0.01 (n/a)245.70 (n/a)143.44 (n/a)113.00 (n/a)99.20 (n/a)61.90 (n/a)

eltwise_add_1_cols_2_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.09 (-45.96%)0.07 (-50.19%)0.06 (-58.92%)0.05 (-35.79%)0.01 (-57.98%)227.00 (+55.80%)194.08 (+95.49%)203.10 (+143.53%)143.50 (+84.92%)35.32 (+23.08%)
897d04e — 2026-03-06 22:12:250.16 (n/a)0.13 (n/a)0.15 (n/a)0.08 (n/a)0.03 (n/a)145.70 (n/a)99.28 (n/a)83.40 (n/a)77.60 (n/a)28.70 (n/a)

eltwise_add_2_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.09 (-23.92%)0.07 (-28.69%)0.07 (-39.52%)0.06 (-19.61%)0.01 (-46.14%)196.70 (+24.41%)168.48 (+37.51%)173.60 (+65.33%)133.30 (+31.46%)24.44 (-11.22%)
897d04e — 2026-03-06 22:12:250.12 (n/a)0.10 (n/a)0.12 (n/a)0.08 (n/a)0.02 (n/a)158.10 (n/a)122.52 (n/a)105.00 (n/a)101.40 (n/a)27.53 (n/a)

eltwise_add_4_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.08 (-51.57%)0.06 (-48.30%)0.06 (-57.53%)0.05 (-19.42%)0.01 (-75.73%)246.50 (+24.06%)202.18 (+74.68%)203.60 (+135.38%)163.80 (+106.56%)31.88 (-37.12%)
897d04e — 2026-03-06 22:12:250.15 (n/a)0.12 (n/a)0.14 (n/a)0.06 (n/a)0.04 (n/a)198.70 (n/a)115.74 (n/a)86.50 (n/a)79.30 (n/a)50.70 (n/a)

eltwise_add_8_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.06 (-43.83%)0.05 (-50.66%)0.05 (-54.13%)0.03 (-57.71%)0.01 (-27.95%)416.30 (+136.40%)263.26 (+110.37%)245.80 (+118.10%)198.40 (+78.10%)88.48 (+210.43%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.10 (n/a)0.11 (n/a)0.07 (n/a)0.02 (n/a)176.10 (n/a)125.14 (n/a)112.70 (n/a)111.40 (n/a)28.50 (n/a)

eltwise_mul_1_cols_2_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.07 (-52.36%)0.06 (-54.68%)0.06 (-58.23%)0.05 (-32.11%)0.01 (-74.51%)244.60 (+47.26%)214.56 (+108.96%)213.60 (+139.46%)175.70 (+109.92%)27.37 (-23.06%)
897d04e — 2026-03-06 22:12:250.15 (n/a)0.13 (n/a)0.14 (n/a)0.07 (n/a)0.03 (n/a)166.10 (n/a)102.68 (n/a)89.20 (n/a)83.70 (n/a)35.57 (n/a)

eltwise_mul_2_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.07 (-51.88%)0.07 (-51.00%)0.07 (-53.62%)0.05 (-43.89%)0.01 (-67.41%)239.60 (+78.27%)191.78 (+99.60%)177.60 (+115.53%)169.30 (+107.73%)28.41 (+24.40%)
897d04e — 2026-03-06 22:12:250.15 (n/a)0.13 (n/a)0.15 (n/a)0.09 (n/a)0.03 (n/a)134.40 (n/a)96.08 (n/a)82.40 (n/a)81.50 (n/a)22.84 (n/a)

eltwise_mul_4_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.07 (-37.59%)0.06 (-42.76%)0.06 (-44.28%)0.04 (-37.16%)0.01 (-45.14%)312.60 (+59.16%)221.98 (+72.61%)207.20 (+79.55%)169.80 (+60.34%)54.30 (+41.68%)
897d04e — 2026-03-06 22:12:250.12 (n/a)0.10 (n/a)0.11 (n/a)0.06 (n/a)0.02 (n/a)196.40 (n/a)128.60 (n/a)115.40 (n/a)105.90 (n/a)38.33 (n/a)

eltwise_mul_8_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.06 (-41.08%)0.05 (-48.87%)0.05 (-52.87%)0.04 (-46.58%)0.01 (-37.07%)290.60 (+87.24%)242.08 (+96.27%)241.30 (+112.23%)190.50 (+69.79%)36.03 (+96.29%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.10 (n/a)0.11 (n/a)0.08 (n/a)0.01 (n/a)155.20 (n/a)123.34 (n/a)113.70 (n/a)112.20 (n/a)18.35 (n/a)

gelu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.07 (-36.79%)0.05 (-45.00%)0.05 (-50.31%)0.04 (-32.38%)0.01 (-40.37%)199.60 (+47.96%)171.34 (+80.59%)176.90 (+101.25%)122.80 (+58.25%)32.12 (+37.86%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.09 (n/a)0.06 (n/a)0.02 (n/a)134.90 (n/a)94.88 (n/a)87.90 (n/a)77.60 (n/a)23.30 (n/a)

gelu_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.04 (-61.39%)0.04 (-62.14%)0.04 (-64.13%)0.03 (-60.80%)0.01 (-64.43%)284.20 (+155.12%)222.70 (+163.24%)215.30 (+178.89%)188.40 (+159.15%)36.86 (+135.41%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.10 (n/a)0.11 (n/a)0.07 (n/a)0.02 (n/a)111.40 (n/a)84.60 (n/a)77.20 (n/a)72.70 (n/a)15.66 (n/a)

gelu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.05 (-52.92%)0.04 (-54.88%)0.04 (-56.69%)0.03 (-46.93%)0.01 (-61.23%)293.70 (+88.39%)204.68 (+115.23%)183.30 (+130.86%)162.80 (+112.26%)52.93 (+54.85%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.10 (n/a)0.05 (n/a)0.02 (n/a)155.90 (n/a)95.10 (n/a)79.40 (n/a)76.70 (n/a)34.18 (n/a)

gelu_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.07 (-33.86%)0.05 (-49.76%)0.04 (-62.31%)0.03 (-38.22%)0.02 (-30.35%)245.60 (+61.90%)186.34 (+101.14%)203.10 (+165.14%)113.20 (+51.13%)55.63 (+67.38%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.10 (n/a)0.11 (n/a)0.05 (n/a)0.02 (n/a)151.70 (n/a)92.64 (n/a)76.60 (n/a)74.90 (n/a)33.24 (n/a)

gelu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.06 (-48.73%)0.04 (-54.99%)0.04 (-55.46%)0.02 (-57.24%)0.01 (-46.78%)353.80 (+133.84%)215.62 (+126.59%)188.10 (+124.46%)145.70 (+95.05%)80.38 (+151.43%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.10 (n/a)0.05 (n/a)0.02 (n/a)151.30 (n/a)95.16 (n/a)83.80 (n/a)74.70 (n/a)31.97 (n/a)

gelu_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.05 (-46.07%)0.05 (-47.65%)0.04 (-53.47%)0.04 (-9.91%)0.01 (-78.19%)201.50 (+11.02%)183.50 (+75.77%)182.40 (+114.84%)153.80 (+85.52%)18.93 (-56.14%)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.09 (n/a)0.10 (n/a)0.05 (n/a)0.02 (n/a)181.50 (n/a)104.40 (n/a)84.90 (n/a)82.90 (n/a)43.16 (n/a)

gelu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.06 (-44.46%)0.05 (-45.06%)0.05 (-46.57%)0.03 (-40.20%)0.01 (-53.36%)239.60 (+67.20%)179.20 (+79.60%)170.90 (+87.19%)146.90 (+80.02%)35.23 (+41.14%)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.09 (n/a)0.09 (n/a)0.06 (n/a)0.02 (n/a)143.30 (n/a)99.78 (n/a)91.30 (n/a)81.60 (n/a)24.96 (n/a)

gelu_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.04 (-53.04%)0.04 (-50.32%)0.04 (-52.91%)0.03 (-27.89%)0.00 (-85.70%)240.80 (+38.63%)225.40 (+92.85%)225.50 (+112.34%)209.10 (+112.93%)13.18 (-58.83%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.07 (n/a)0.08 (n/a)0.05 (n/a)0.01 (n/a)173.70 (n/a)116.88 (n/a)106.20 (n/a)98.20 (n/a)32.02 (n/a)

gemm_1792x896x1152_64x32x48_8cols_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
27cd2f3 — 2026-03-20 17:09:574.85 (-2.85%)4.44 (-9.09%)4.50 (-9.48%)3.94 (-12.68%)0.42 (+103.76%)2386.10 (+14.52%)2133.76 (+10.64%)2091.90 (+10.47%)1937.40 (+2.93%)206.31 (+137.76%)1909.46 (-2.85%)1746.59 (-9.09%)1768.44 (-9.48%)1550.39 (-12.68%)166.21 (+103.76%)
897d04e — 2026-03-06 22:12:255.00 (n/a)4.88 (n/a)4.97 (n/a)4.51 (n/a)0.21 (n/a)2083.60 (n/a)1928.60 (n/a)1893.70 (n/a)1882.30 (n/a)86.77 (n/a)1965.38 (n/a)1921.12 (n/a)1953.55 (n/a)1775.47 (n/a)81.57 (n/a)

gemm_192x384x64_48x96x16_4cols

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
27cd2f3 — 2026-03-20 17:09:571.01 (-50.52%)0.73 (-55.71%)0.61 (-67.35%)0.54 (-31.42%)0.21 (-59.35%)410.60 (+45.81%)320.84 (+111.47%)363.30 (+206.32%)218.20 (+102.22%)85.28 (+15.39%)43.26 (-50.52%)31.35 (-55.71%)25.97 (-67.35%)22.99 (-31.42%)9.14 (-59.35%)
897d04e — 2026-03-06 22:12:252.05 (n/a)1.66 (n/a)1.86 (n/a)0.79 (n/a)0.53 (n/a)281.60 (n/a)151.72 (n/a)118.60 (n/a)107.90 (n/a)73.91 (n/a)87.43 (n/a)70.78 (n/a)79.55 (n/a)33.52 (n/a)22.49 (n/a)

gemm_192x384x64_48x96x16_4cols_bcolmaj_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
27cd2f3 — 2026-03-20 17:09:571.67 (-25.39%)1.10 (-39.71%)1.03 (-45.17%)0.76 (-42.25%)0.34 (+3.23%)290.60 (+73.18%)214.38 (+71.75%)215.80 (+82.42%)132.80 (+34.01%)56.32 (+120.21%)71.06 (-25.39%)46.94 (-39.71%)43.74 (-45.17%)32.47 (-42.25%)14.41 (+3.23%)
897d04e — 2026-03-06 22:12:252.23 (n/a)1.82 (n/a)1.87 (n/a)1.32 (n/a)0.33 (n/a)167.80 (n/a)124.82 (n/a)118.30 (n/a)99.10 (n/a)25.58 (n/a)95.25 (n/a)77.86 (n/a)79.77 (n/a)56.23 (n/a)13.96 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.72 (-0.11%)3.55 (-2.75%)3.56 (-2.75%)3.36 (-5.33%)0.12 (+76.10%)2725.23 (-0.11%)2602.52 (-2.75%)2612.11 (-2.75%)2459.89 (-5.33%)86.27 (+76.10%)6984.00 (+5.63%)6608.60 (+2.91%)6577.00 (+2.83%)6304.00 (+0.11%)221.87 (+87.15%)
9e98049 — 2025-11-25 20:50:553.72 (n/a)3.65 (n/a)3.66 (n/a)3.54 (n/a)0.07 (n/a)2728.26 (n/a)2676.06 (n/a)2686.03 (n/a)2598.29 (n/a)48.99 (n/a)6612.00 (n/a)6422.00 (n/a)6396.00 (n/a)6297.00 (n/a)118.55 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_0_bcolmaj_1_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.40 (-3.14%)3.38 (-0.82%)3.38 (-0.82%)3.35 (+4.93%)0.02 (-84.51%)2493.81 (-3.14%)2475.27 (-0.82%)2479.06 (-0.82%)2453.92 (+4.93%)13.33 (-84.51%)7001.00 (-4.70%)6940.80 (+0.71%)6930.00 (+0.83%)6889.00 (+3.24%)37.43 (-84.76%)
9e98049 — 2025-11-25 20:50:553.51 (n/a)3.40 (n/a)3.41 (n/a)3.19 (n/a)0.12 (n/a)2574.53 (n/a)2495.72 (n/a)2499.62 (n/a)2338.67 (n/a)86.02 (n/a)7346.00 (n/a)6892.20 (n/a)6873.00 (n/a)6673.00 (n/a)245.57 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_1_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.02 (+0.52%)2.87 (-2.05%)2.86 (-2.24%)2.70 (-6.16%)0.11 (+131.66%)2216.19 (+0.52%)2104.44 (-2.05%)2094.08 (-2.24%)1980.84 (-6.16%)80.25 (+131.67%)8673.00 (+6.56%)8175.60 (+2.21%)8204.00 (+2.29%)7752.00 (-0.51%)314.03 (+145.36%)
9e98049 — 2025-11-25 20:50:553.01 (n/a)2.93 (n/a)2.92 (n/a)2.88 (n/a)0.05 (n/a)2204.81 (n/a)2148.41 (n/a)2142.13 (n/a)2110.81 (n/a)34.64 (n/a)8139.00 (n/a)7998.60 (n/a)8020.00 (n/a)7792.00 (n/a)127.99 (n/a)

gemm_2048x2048x2048_64x64x64_1cols

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
27cd2f3 — 2026-03-20 17:09:570.51 (-1.82%)0.51 (-1.96%)0.51 (-1.86%)0.51 (-2.27%)0.00 (+30.18%)49527.20 (+2.33%)49284.16 (+2.00%)49307.10 (+1.89%)48904.40 (+1.86%)231.55 (+35.57%)351.29 (-1.82%)348.59 (-1.96%)348.43 (-1.86%)346.88 (-2.27%)1.64 (+30.17%)
897d04e — 2026-03-06 22:12:250.52 (n/a)0.52 (n/a)0.52 (n/a)0.52 (n/a)0.00 (n/a)48401.80 (n/a)48318.62 (n/a)48391.10 (n/a)48013.20 (n/a)170.80 (n/a)357.82 (n/a)355.56 (n/a)355.02 (n/a)354.94 (n/a)1.26 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:491.04 (+0.94%)1.03 (+0.23%)1.03 (+0.09%)1.02 (+0.07%)0.00 (+197.51%)24576.00 (-0.07%)24485.82 (-0.23%)24530.40 (-0.09%)24266.40 (-0.94%)111.26 (+194.35%)
0a6c11c — 2025-12-04 00:39:101.03 (n/a)1.03 (n/a)1.03 (n/a)1.02 (n/a)0.00 (n/a)24594.30 (n/a)24542.14 (n/a)24551.90 (n/a)24495.50 (n/a)37.80 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:051.04 (+0.28%)1.03 (-0.15%)1.03 (+0.05%)1.02 (-0.52%)0.01 (+42.72%)24719.80 (+0.52%)24512.42 (+0.16%)24529.20 (-0.05%)24092.60 (-0.28%)252.92 (+43.15%)713.08 (+0.28%)700.92 (-0.15%)700.38 (+0.05%)694.98 (-0.52%)7.30 (+42.71%)
50ad37c — 2025-12-29 20:35:531.04 (n/a)1.03 (n/a)1.03 (n/a)1.02 (n/a)0.01 (n/a)24592.30 (n/a)24473.70 (n/a)24540.90 (n/a)24160.80 (n/a)176.68 (n/a)711.06 (n/a)702.00 (n/a)700.05 (n/a)698.59 (n/a)5.12 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_1_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:490.99 (+1.69%)0.97 (+0.29%)0.97 (-0.04%)0.97 (-0.15%)0.01 (+434.92%)26025.30 (+0.15%)25839.34 (-0.29%)25916.70 (+0.04%)25446.00 (-1.67%)202.05 (+425.44%)
0a6c11c — 2025-12-04 00:39:100.97 (n/a)0.97 (n/a)0.97 (n/a)0.97 (n/a)0.00 (n/a)25986.90 (n/a)25913.52 (n/a)25905.70 (n/a)25877.00 (n/a)38.45 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_1_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:050.97 (-0.01%)0.97 (+0.03%)0.97 (+0.04%)0.97 (+0.00%)0.00 (-6.43%)25927.90 (-0.00%)25903.20 (-0.03%)25898.60 (-0.04%)25892.80 (+0.01%)14.05 (-6.56%)663.50 (-0.01%)663.23 (+0.03%)663.35 (+0.04%)662.60 (+0.00%)0.36 (-6.43%)
50ad37c — 2025-12-29 20:35:530.97 (n/a)0.97 (n/a)0.97 (n/a)0.97 (n/a)0.00 (n/a)25928.10 (n/a)25910.14 (n/a)25908.10 (n/a)25891.20 (n/a)15.04 (n/a)663.54 (n/a)663.06 (n/a)663.11 (n/a)662.60 (n/a)0.38 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_1_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:490.92 (-1.02%)0.92 (-0.66%)0.92 (-0.10%)0.91 (-0.78%)0.01 (-4.54%)27751.40 (+0.79%)27480.74 (+0.66%)27354.80 (+0.10%)27320.50 (+1.03%)182.69 (-2.84%)
0a6c11c — 2025-12-04 00:39:100.93 (n/a)0.92 (n/a)0.92 (n/a)0.91 (n/a)0.01 (n/a)27534.80 (n/a)27299.32 (n/a)27326.70 (n/a)27042.30 (n/a)188.04 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_1_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:050.92 (+0.11%)0.91 (-0.12%)0.91 (-0.21%)0.91 (-0.24%)0.00 (+30.34%)27723.50 (+0.24%)27548.94 (+0.12%)27522.90 (+0.21%)27403.60 (-0.11%)124.26 (+30.58%)626.92 (+0.11%)623.62 (-0.12%)624.20 (-0.21%)619.69 (-0.24%)2.81 (+30.34%)
50ad37c — 2025-12-29 20:35:530.92 (n/a)0.91 (n/a)0.92 (n/a)0.91 (n/a)0.00 (n/a)27657.10 (n/a)27515.58 (n/a)27464.70 (n/a)27434.60 (n/a)95.16 (n/a)626.21 (n/a)624.37 (n/a)625.52 (n/a)621.17 (n/a)2.16 (n/a)

gemm_2048x2048x2048_64x64x64_2cols_bcolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
27cd2f3 — 2026-03-20 17:09:570.89 (-2.16%)0.88 (-2.69%)0.87 (-3.78%)0.87 (-1.99%)0.01 (+1.27%)29024.80 (+2.03%)28669.54 (+2.76%)28864.10 (+3.92%)28127.40 (+2.21%)379.38 (+5.49%)610.79 (-2.16%)599.32 (-2.69%)595.20 (-3.78%)591.90 (-1.99%)7.98 (+1.27%)
897d04e — 2026-03-06 22:12:250.91 (n/a)0.90 (n/a)0.91 (n/a)0.88 (n/a)0.01 (n/a)28447.10 (n/a)27899.18 (n/a)27774.40 (n/a)27519.50 (n/a)359.64 (n/a)624.28 (n/a)615.87 (n/a)618.55 (n/a)603.92 (n/a)7.88 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
31ab24e — 2025-11-22 02:01:563.65 (-1.06%)3.59 (+0.83%)3.63 (-0.26%)3.45 (+2.66%)0.07 (-44.71%)2678.08 (-1.06%)2632.13 (+0.83%)2658.60 (-0.26%)2528.31 (+2.66%)53.86 (-44.71%)6795.00 (-2.59%)6529.80 (-0.92%)6462.00 (+0.26%)6415.00 (+1.07%)137.14 (-45.27%)
d9a64c5 — 2025-11-21 20:49:353.69 (n/a)3.56 (n/a)3.64 (n/a)3.36 (n/a)0.13 (n/a)2706.77 (n/a)2610.50 (n/a)2665.61 (n/a)2462.71 (n/a)97.41 (n/a)6976.00 (n/a)6590.40 (n/a)6445.00 (n/a)6347.00 (n/a)250.58 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.98 (+0.20%)3.84 (-2.95%)3.76 (-5.27%)3.72 (-4.97%)0.13 (+473.97%)6760.00 (+5.23%)6559.50 (+3.13%)6697.10 (+5.56%)6321.00 (-0.20%)218.68 (+499.40%)2717.91 (+0.20%)2621.44 (-2.95%)2565.27 (-5.27%)2541.40 (-4.97%)88.37 (+473.97%)
331dcca — 2026-01-08 18:34:103.97 (n/a)3.96 (n/a)3.97 (n/a)3.92 (n/a)0.02 (n/a)6423.80 (n/a)6360.32 (n/a)6344.30 (n/a)6333.90 (n/a)36.48 (n/a)2712.39 (n/a)2701.17 (n/a)2707.91 (n/a)2674.40 (n/a)15.40 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_1_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.75 (+0.28%)3.70 (+1.25%)3.71 (-0.21%)3.65 (+4.77%)0.04 (-66.90%)6901.80 (-4.56%)6795.46 (-1.30%)6790.20 (+0.21%)6715.90 (-0.28%)67.05 (-68.40%)2558.08 (+0.28%)2528.33 (+1.25%)2530.10 (-0.21%)2489.18 (+4.77%)24.81 (-66.90%)
331dcca — 2026-01-08 18:34:103.74 (n/a)3.66 (n/a)3.71 (n/a)3.48 (n/a)0.11 (n/a)7231.20 (n/a)6885.18 (n/a)6775.70 (n/a)6734.90 (n/a)212.21 (n/a)2550.87 (n/a)2497.04 (n/a)2535.50 (n/a)2375.78 (n/a)74.96 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_1_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.62 (-0.77%)3.54 (+0.14%)3.60 (-0.04%)3.44 (+2.17%)0.09 (-25.77%)7307.40 (-2.13%)7103.46 (-0.18%)6982.50 (+0.04%)6948.20 (+0.77%)182.95 (-26.55%)2472.55 (-0.77%)2419.80 (+0.14%)2460.41 (-0.04%)2351.04 (+2.17%)61.76 (-25.77%)
331dcca — 2026-01-08 18:34:103.65 (n/a)3.54 (n/a)3.61 (n/a)3.37 (n/a)0.12 (n/a)7466.20 (n/a)7116.40 (n/a)6979.60 (n/a)6895.10 (n/a)249.09 (n/a)2491.61 (n/a)2416.45 (n/a)2461.45 (n/a)2301.00 (n/a)83.20 (n/a)

gemm_2048x2048x2048_64x64x64_8cols_bcolmaj_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
27cd2f3 — 2026-03-20 17:09:573.45 (-1.80%)3.36 (-3.53%)3.41 (-1.93%)3.27 (-6.08%)0.09 (+618.35%)7704.00 (+6.48%)7483.34 (+3.72%)7369.30 (+1.97%)7303.00 (+1.83%)199.39 (+681.51%)2352.44 (-1.80%)2297.04 (-3.53%)2331.28 (-1.93%)2229.99 (-6.08%)60.68 (+618.34%)
897d04e — 2026-03-06 22:12:253.51 (n/a)3.49 (n/a)3.48 (n/a)3.48 (n/a)0.01 (n/a)7235.40 (n/a)7215.10 (n/a)7227.20 (n/a)7171.80 (n/a)25.51 (n/a)2395.46 (n/a)2381.12 (n/a)2377.13 (n/a)2374.41 (n/a)8.45 (n/a)

gemm_384x1536x1792_32x48x64_4cols_bcolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
27cd2f3 — 2026-03-20 17:09:574.02 (-7.52%)3.54 (-14.96%)3.67 (-15.13%)3.15 (-10.62%)0.37 (+2.91%)2555.40 (+11.88%)2295.02 (+17.83%)2199.20 (+17.82%)2007.00 (+8.12%)237.95 (+26.56%)1053.27 (-7.52%)929.05 (-14.96%)961.24 (-15.13%)827.25 (-10.62%)96.11 (+2.91%)
897d04e — 2026-03-06 22:12:254.34 (n/a)4.17 (n/a)4.32 (n/a)3.53 (n/a)0.36 (n/a)2284.00 (n/a)1947.78 (n/a)1866.50 (n/a)1856.20 (n/a)188.02 (n/a)1138.86 (n/a)1092.52 (n/a)1132.55 (n/a)925.53 (n/a)93.39 (n/a)

gemm_896x1792x640_32x64x80_8cols_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
27cd2f3 — 2026-03-20 17:09:575.09 (-24.31%)4.31 (-31.70%)4.78 (-27.67%)3.21 (-38.20%)0.85 (+31.66%)2072.60 (+61.82%)1598.22 (+50.13%)1390.50 (+38.25%)1307.20 (+32.12%)344.71 (+178.34%)1572.16 (-24.31%)1331.17 (-31.70%)1478.08 (-27.67%)991.62 (-38.20%)262.95 (+31.66%)
897d04e — 2026-03-06 22:12:256.72 (n/a)6.31 (n/a)6.61 (n/a)5.19 (n/a)0.65 (n/a)1280.80 (n/a)1064.58 (n/a)1005.80 (n/a)989.40 (n/a)123.84 (n/a)2077.17 (n/a)1949.09 (n/a)2043.44 (n/a)1604.60 (n/a)199.72 (n/a)

layer_norm_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.05 (-48.93%)0.05 (-44.12%)0.04 (-49.75%)0.04 (-25.27%)0.01 (-74.05%)200.30 (+33.80%)178.76 (+69.73%)188.40 (+98.94%)153.50 (+95.79%)21.57 (-30.67%)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)149.70 (n/a)105.32 (n/a)94.70 (n/a)78.40 (n/a)31.11 (n/a)

layer_norm_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.07 (-38.78%)0.05 (-48.63%)0.04 (-58.80%)0.04 (-32.73%)0.01 (-50.47%)207.70 (+48.68%)180.34 (+89.95%)188.80 (+142.67%)122.30 (+63.28%)33.47 (+18.58%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.11 (n/a)0.06 (n/a)0.02 (n/a)139.70 (n/a)94.94 (n/a)77.80 (n/a)74.90 (n/a)28.23 (n/a)

layer_norm_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.07 (-33.77%)0.05 (-41.62%)0.05 (-50.45%)0.04 (-40.55%)0.02 (-10.09%)229.10 (+68.21%)169.04 (+77.30%)172.40 (+101.87%)117.50 (+51.03%)50.07 (+111.85%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.10 (n/a)0.06 (n/a)0.02 (n/a)136.20 (n/a)95.34 (n/a)85.40 (n/a)77.80 (n/a)23.63 (n/a)

layer_norm_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.06 (-45.40%)0.05 (-45.44%)0.04 (-53.45%)0.03 (-23.84%)0.01 (-52.18%)282.30 (+31.30%)194.44 (+71.31%)203.10 (+114.92%)139.20 (+83.16%)58.35 (+1.75%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.08 (n/a)0.09 (n/a)0.04 (n/a)0.03 (n/a)215.00 (n/a)113.50 (n/a)94.50 (n/a)76.00 (n/a)57.35 (n/a)

layer_norm_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.07 (-37.46%)0.05 (-43.20%)0.04 (-53.67%)0.04 (-16.92%)0.01 (-43.96%)228.20 (+20.42%)179.96 (+68.50%)195.70 (+115.77%)124.70 (+59.87%)46.53 (+0.01%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.09 (n/a)0.04 (n/a)0.02 (n/a)189.50 (n/a)106.80 (n/a)90.70 (n/a)78.00 (n/a)46.53 (n/a)

layer_norm_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.06 (-41.57%)0.04 (-52.99%)0.03 (-61.59%)0.03 (-32.15%)0.01 (-48.25%)264.40 (+47.38%)226.88 (+106.89%)244.10 (+160.23%)146.20 (+71.19%)46.35 (+18.10%)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)179.40 (n/a)109.66 (n/a)93.80 (n/a)85.40 (n/a)39.25 (n/a)

layer_norm_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.06 (-38.94%)0.05 (-39.23%)0.05 (-40.10%)0.03 (-35.07%)0.01 (-32.47%)242.20 (+53.97%)187.36 (+65.78%)169.30 (+66.96%)139.00 (+63.72%)47.81 (+71.77%)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.08 (n/a)0.08 (n/a)0.05 (n/a)0.02 (n/a)157.30 (n/a)113.02 (n/a)101.40 (n/a)84.90 (n/a)27.83 (n/a)

layer_norm_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.04 (-49.53%)0.04 (-47.51%)0.04 (-54.17%)0.03 (-20.76%)0.00 (-80.40%)249.90 (+26.21%)216.54 (+78.13%)215.50 (+118.12%)191.50 (+98.24%)21.32 (-51.00%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)198.00 (n/a)121.56 (n/a)98.80 (n/a)96.60 (n/a)43.50 (n/a)

matrix_vector_mul_128x128_32_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:490.37 (+17.13%)0.30 (+24.87%)0.30 (+38.03%)0.23 (+12.47%)0.06 (+35.23%)
0a6c11c — 2025-12-04 00:39:100.32 (n/a)0.24 (n/a)0.22 (n/a)0.20 (n/a)0.04 (n/a)

matrix_vector_mul_128x128_32_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:100.26 (-1.81%)0.24 (+4.06%)0.24 (+0.52%)0.21 (+16.97%)0.02 (-48.16%)0.26 (-1.81%)0.24 (+4.06%)0.24 (+0.52%)0.20 (+16.97%)0.02 (-48.16%)
a4b6ffe — 2026-01-05 19:24:050.26 (n/a)0.23 (n/a)0.24 (n/a)0.18 (n/a)0.04 (n/a)0.26 (n/a)0.23 (n/a)0.24 (n/a)0.17 (n/a)0.04 (n/a)

matrix_vector_mul_128x128_32tsi_128tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
27cd2f3 — 2026-03-20 17:09:570.23 (n/a)0.18 (n/a)0.17 (n/a)0.13 (n/a)0.04 (n/a)0.23 (n/a)0.18 (n/a)0.17 (n/a)0.13 (n/a)0.04 (n/a)

matrix_vector_mul_128x128_32tsi_128tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:250.37 (+5.44%)0.31 (-0.05%)0.34 (-1.99%)0.22 (-7.87%)0.06 (+15.13%)0.36 (+5.44%)0.31 (-0.05%)0.34 (-1.99%)0.21 (-7.87%)0.06 (+15.13%)
84d3478 — 2026-02-17 23:25:130.35 (n/a)0.31 (n/a)0.35 (n/a)0.23 (n/a)0.05 (n/a)0.35 (n/a)0.31 (n/a)0.34 (n/a)0.23 (n/a)0.05 (n/a)

matrix_vector_mul_2048x8192_1_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4913.57 (+3.47%)13.15 (+1.09%)13.10 (+0.16%)12.52 (-1.58%)0.37 (+153.19%)
0a6c11c — 2025-12-04 00:39:1013.11 (n/a)13.01 (n/a)13.07 (n/a)12.72 (n/a)0.15 (n/a)

matrix_vector_mul_2048x8192_1_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1013.43 (-0.07%)13.30 (+4.62%)13.27 (+1.29%)13.16 (+16.70%)0.12 (-86.69%)13.42 (-0.07%)13.29 (+4.62%)13.26 (+1.29%)13.16 (+16.70%)0.12 (-86.69%)
a4b6ffe — 2026-01-05 19:24:0513.44 (n/a)12.71 (n/a)13.10 (n/a)11.28 (n/a)0.87 (n/a)13.43 (n/a)12.70 (n/a)13.09 (n/a)11.27 (n/a)0.87 (n/a)

matrix_vector_mul_2048x8192_1_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4925.06 (+2.47%)23.71 (+1.81%)24.13 (+0.50%)20.90 (+5.41%)1.48 (-15.60%)
0a6c11c — 2025-12-04 00:39:1024.46 (n/a)23.29 (n/a)24.01 (n/a)19.83 (n/a)1.75 (n/a)

matrix_vector_mul_2048x8192_1_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1024.46 (-0.60%)21.07 (-8.03%)19.75 (-17.83%)18.08 (+1.18%)3.03 (+7.07%)24.44 (-0.60%)21.06 (-8.03%)19.74 (-17.83%)18.07 (+1.18%)3.03 (+7.07%)
a4b6ffe — 2026-01-05 19:24:0524.60 (n/a)22.91 (n/a)24.04 (n/a)17.87 (n/a)2.83 (n/a)24.59 (n/a)22.90 (n/a)24.02 (n/a)17.86 (n/a)2.83 (n/a)

matrix_vector_mul_2048x8192_1_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4942.05 (+1.54%)40.50 (+12.60%)40.49 (+2.36%)38.98 (+97.53%)0.97 (-88.09%)
0a6c11c — 2025-12-04 00:39:1041.41 (n/a)35.97 (n/a)39.55 (n/a)19.73 (n/a)8.15 (n/a)

matrix_vector_mul_2048x8192_1_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1042.58 (+1.51%)41.37 (+14.29%)41.25 (+4.89%)40.33 (+91.42%)0.92 (-89.18%)42.56 (+1.51%)41.34 (+14.29%)41.23 (+4.89%)40.30 (+91.42%)0.92 (-89.18%)
a4b6ffe — 2026-01-05 19:24:0541.95 (n/a)36.20 (n/a)39.33 (n/a)21.07 (n/a)8.54 (n/a)41.92 (n/a)36.17 (n/a)39.31 (n/a)21.05 (n/a)8.53 (n/a)

matrix_vector_mul_2048x8192_1_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4946.62 (+3.11%)44.82 (+0.17%)44.78 (+0.17%)42.60 (-3.80%)1.60 (+367.72%)
0a6c11c — 2025-12-04 00:39:1045.21 (n/a)44.75 (n/a)44.70 (n/a)44.29 (n/a)0.34 (n/a)

matrix_vector_mul_2048x8192_1_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1046.04 (+4.14%)45.06 (+3.47%)44.55 (+2.76%)44.30 (+2.84%)0.87 (+66.58%)46.01 (+4.14%)45.03 (+3.47%)44.52 (+2.76%)44.27 (+2.84%)0.87 (+66.58%)
a4b6ffe — 2026-01-05 19:24:0544.21 (n/a)43.55 (n/a)43.35 (n/a)43.07 (n/a)0.52 (n/a)44.18 (n/a)43.52 (n/a)43.32 (n/a)43.05 (n/a)0.52 (n/a)

matrix_vector_mul_2048x8192_1tsi_1024tso_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
27cd2f3 — 2026-03-20 17:09:5724.03 (n/a)23.27 (n/a)22.95 (n/a)22.68 (n/a)0.58 (n/a)24.02 (n/a)23.25 (n/a)22.93 (n/a)22.67 (n/a)0.58 (n/a)

matrix_vector_mul_2048x8192_1tsi_1024tso_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2524.47 (-3.32%)24.16 (-3.13%)24.20 (-3.09%)23.84 (-3.04%)0.25 (-14.36%)24.46 (-3.32%)24.15 (-3.13%)24.18 (-3.09%)23.83 (-3.04%)0.25 (-14.36%)
84d3478 — 2026-02-17 23:25:1325.31 (n/a)24.94 (n/a)24.97 (n/a)24.59 (n/a)0.29 (n/a)25.30 (n/a)24.93 (n/a)24.95 (n/a)24.58 (n/a)0.29 (n/a)

matrix_vector_mul_2048x8192_1tsi_2048tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
27cd2f3 — 2026-03-20 17:09:5713.33 (n/a)12.98 (n/a)13.15 (n/a)11.99 (n/a)0.56 (n/a)13.32 (n/a)12.97 (n/a)13.14 (n/a)11.98 (n/a)0.56 (n/a)

matrix_vector_mul_2048x8192_1tsi_2048tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2513.25 (-3.04%)13.02 (-2.92%)13.10 (-2.64%)12.56 (-4.14%)0.27 (+32.55%)13.24 (-3.04%)13.01 (-2.92%)13.09 (-2.64%)12.55 (-4.14%)0.27 (+32.55%)
84d3478 — 2026-02-17 23:25:1313.66 (n/a)13.41 (n/a)13.45 (n/a)13.10 (n/a)0.20 (n/a)13.65 (n/a)13.40 (n/a)13.44 (n/a)13.09 (n/a)0.20 (n/a)

matrix_vector_mul_2048x8192_1tsi_256tso_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
27cd2f3 — 2026-03-20 17:09:5744.75 (n/a)40.49 (n/a)39.29 (n/a)38.23 (n/a)2.70 (n/a)44.72 (n/a)40.47 (n/a)39.27 (n/a)38.20 (n/a)2.70 (n/a)

matrix_vector_mul_2048x8192_1tsi_256tso_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2547.71 (+3.51%)45.38 (+9.55%)45.22 (-1.27%)43.47 (+75.24%)1.52 (-83.67%)47.68 (+3.51%)45.35 (+9.55%)45.19 (-1.27%)43.45 (+75.24%)1.52 (-83.67%)
84d3478 — 2026-02-17 23:25:1346.09 (n/a)41.43 (n/a)45.80 (n/a)24.81 (n/a)9.31 (n/a)46.06 (n/a)41.40 (n/a)45.77 (n/a)24.79 (n/a)9.30 (n/a)

matrix_vector_mul_2048x8192_1tsi_512tso_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
27cd2f3 — 2026-03-20 17:09:5739.27 (n/a)38.51 (n/a)39.22 (n/a)37.13 (n/a)1.03 (n/a)39.25 (n/a)38.49 (n/a)39.19 (n/a)37.10 (n/a)1.03 (n/a)

matrix_vector_mul_2048x8192_1tsi_512tso_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2540.84 (-0.32%)39.53 (-1.25%)39.40 (-1.97%)38.87 (-0.69%)0.77 (-6.24%)40.81 (-0.32%)39.50 (-1.25%)39.37 (-1.97%)38.84 (-0.69%)0.77 (-6.24%)
84d3478 — 2026-02-17 23:25:1340.97 (n/a)40.03 (n/a)40.19 (n/a)39.14 (n/a)0.82 (n/a)40.95 (n/a)40.00 (n/a)40.17 (n/a)39.11 (n/a)0.82 (n/a)

matrix_vector_mul_8192x2048_4_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4913.29 (+1.63%)13.19 (+6.08%)13.19 (+1.17%)13.09 (+26.86%)0.08 (-92.94%)
0a6c11c — 2025-12-04 00:39:1013.08 (n/a)12.44 (n/a)13.03 (n/a)10.32 (n/a)1.07 (n/a)

matrix_vector_mul_8192x2048_4_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1013.68 (+3.42%)13.30 (+8.54%)13.38 (+7.25%)12.72 (+14.96%)0.35 (-66.09%)13.67 (+3.42%)13.29 (+8.54%)13.38 (+7.25%)12.71 (+14.96%)0.35 (-66.09%)
a4b6ffe — 2026-01-05 19:24:0513.23 (n/a)12.25 (n/a)12.48 (n/a)11.07 (n/a)1.04 (n/a)13.22 (n/a)12.24 (n/a)12.47 (n/a)11.06 (n/a)1.04 (n/a)

matrix_vector_mul_8192x2048_4_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4925.58 (+4.24%)24.23 (+6.59%)24.03 (+0.29%)23.58 (+37.29%)0.70 (-74.97%)
0a6c11c — 2025-12-04 00:39:1024.54 (n/a)22.73 (n/a)23.96 (n/a)17.18 (n/a)2.79 (n/a)

matrix_vector_mul_8192x2048_4_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1025.08 (+1.32%)24.68 (+13.42%)24.61 (+3.26%)24.47 (+48.32%)0.24 (-93.56%)25.07 (+1.32%)24.67 (+13.42%)24.59 (+3.26%)24.46 (+48.32%)0.24 (-93.56%)
a4b6ffe — 2026-01-05 19:24:0524.76 (n/a)21.76 (n/a)23.83 (n/a)16.50 (n/a)3.74 (n/a)24.74 (n/a)21.75 (n/a)23.82 (n/a)16.49 (n/a)3.73 (n/a)

matrix_vector_mul_8192x2048_4_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4940.94 (-2.19%)39.74 (-1.10%)39.76 (-0.60%)38.17 (-2.14%)1.04 (+4.39%)
0a6c11c — 2025-12-04 00:39:1041.86 (n/a)40.18 (n/a)40.00 (n/a)39.00 (n/a)1.00 (n/a)

matrix_vector_mul_8192x2048_4_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1040.74 (-1.09%)37.16 (-7.15%)40.25 (+1.47%)24.51 (-37.44%)7.07 (+736.13%)40.72 (-1.09%)37.13 (-7.15%)40.23 (+1.47%)24.50 (-37.44%)7.07 (+736.13%)
a4b6ffe — 2026-01-05 19:24:0541.19 (n/a)40.02 (n/a)39.67 (n/a)39.18 (n/a)0.85 (n/a)41.17 (n/a)39.99 (n/a)39.64 (n/a)39.16 (n/a)0.85 (n/a)

matrix_vector_mul_8192x2048_4_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4946.71 (-1.04%)44.91 (-2.34%)44.71 (-2.98%)43.80 (-1.44%)0.96 (+6.93%)
0a6c11c — 2025-12-04 00:39:1047.20 (n/a)45.98 (n/a)46.08 (n/a)44.44 (n/a)0.90 (n/a)

matrix_vector_mul_8192x2048_4_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1046.28 (+5.68%)44.89 (+14.59%)44.74 (+4.26%)44.20 (+62.80%)0.84 (-87.96%)46.25 (+5.68%)44.86 (+14.59%)44.71 (+4.26%)44.17 (+62.80%)0.84 (-87.96%)
a4b6ffe — 2026-01-05 19:24:0543.79 (n/a)39.17 (n/a)42.91 (n/a)27.15 (n/a)6.98 (n/a)43.77 (n/a)39.15 (n/a)42.88 (n/a)27.13 (n/a)6.98 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
27cd2f3 — 2026-03-20 17:09:5713.31 (n/a)13.04 (n/a)13.09 (n/a)12.67 (n/a)0.24 (n/a)13.30 (n/a)13.04 (n/a)13.08 (n/a)12.66 (n/a)0.24 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2513.34 (-0.79%)13.20 (-0.36%)13.18 (-1.55%)13.12 (+1.75%)0.08 (-65.93%)13.33 (-0.79%)13.19 (-0.36%)13.17 (-1.55%)13.11 (+1.75%)0.08 (-65.93%)
84d3478 — 2026-02-17 23:25:1313.45 (n/a)13.25 (n/a)13.39 (n/a)12.90 (n/a)0.25 (n/a)13.44 (n/a)13.24 (n/a)13.38 (n/a)12.89 (n/a)0.25 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
27cd2f3 — 2026-03-20 17:09:5724.92 (n/a)23.97 (n/a)24.01 (n/a)23.16 (n/a)0.65 (n/a)24.90 (n/a)23.95 (n/a)23.99 (n/a)23.15 (n/a)0.65 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2524.10 (-5.45%)23.07 (-1.85%)24.01 (-2.88%)19.26 (+7.47%)2.13 (-32.15%)24.08 (-5.45%)23.05 (-1.85%)23.99 (-2.88%)19.25 (+7.47%)2.13 (-32.15%)
84d3478 — 2026-02-17 23:25:1325.48 (n/a)23.50 (n/a)24.72 (n/a)17.92 (n/a)3.14 (n/a)25.47 (n/a)23.48 (n/a)24.71 (n/a)17.91 (n/a)3.14 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
27cd2f3 — 2026-03-20 17:09:5739.84 (n/a)37.22 (n/a)36.81 (n/a)35.31 (n/a)1.81 (n/a)39.81 (n/a)37.20 (n/a)36.78 (n/a)35.29 (n/a)1.81 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2540.80 (-0.16%)39.58 (-1.93%)39.75 (-1.71%)38.28 (-2.50%)1.10 (+67.33%)40.78 (-0.16%)39.55 (-1.93%)39.73 (-1.71%)38.26 (-2.50%)1.10 (+67.33%)
84d3478 — 2026-02-17 23:25:1340.87 (n/a)40.36 (n/a)40.45 (n/a)39.26 (n/a)0.66 (n/a)40.84 (n/a)40.33 (n/a)40.42 (n/a)39.24 (n/a)0.65 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
27cd2f3 — 2026-03-20 17:09:5743.91 (n/a)41.60 (n/a)41.57 (n/a)39.07 (n/a)1.76 (n/a)43.88 (n/a)41.57 (n/a)41.54 (n/a)39.05 (n/a)1.76 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2549.31 (+6.27%)46.12 (+1.67%)45.30 (-0.27%)44.62 (+1.87%)1.85 (+85.10%)49.28 (+6.27%)46.10 (+1.67%)45.28 (-0.27%)44.59 (+1.87%)1.85 (+85.10%)
84d3478 — 2026-02-17 23:25:1346.40 (n/a)45.37 (n/a)45.43 (n/a)43.80 (n/a)1.00 (n/a)46.37 (n/a)45.34 (n/a)45.40 (n/a)43.78 (n/a)1.00 (n/a)

mem_copy_16_cores_2_chans_2048_tile_128_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.05 (-30.99%)0.04 (-41.89%)0.04 (-42.42%)0.02 (-56.80%)0.01 (+35.99%)401.00 (+131.39%)252.10 (+82.76%)232.00 (+73.65%)178.80 (+44.89%)87.19 (+373.80%)
130b6ea — 2025-12-05 21:15:490.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)173.30 (n/a)137.94 (n/a)133.60 (n/a)123.40 (n/a)18.40 (n/a)

mem_copy_16_cores_2_chans_2048_tile_128_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-2.12%)0.06 (-16.44%)0.06 (-21.48%)0.03 (-23.84%)0.02 (+10.15%)247.60 (+31.28%)148.38 (+24.08%)128.30 (+27.41%)100.10 (+2.14%)58.70 (+50.73%)
84d3478 — 2026-02-17 23:25:130.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)188.60 (n/a)119.58 (n/a)100.70 (n/a)98.00 (n/a)38.94 (n/a)

mem_copy_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (-16.99%)0.05 (-12.47%)0.05 (-7.32%)0.04 (-11.17%)0.00 (-29.61%)197.00 (+12.57%)171.00 (+13.70%)164.00 (+7.89%)153.00 (+20.47%)17.49 (-4.32%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)175.00 (n/a)150.40 (n/a)152.00 (n/a)127.00 (n/a)18.28 (n/a)

mem_copy_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (+5.56%)0.05 (-1.59%)0.05 (-1.39%)0.03 (-14.07%)0.01 (+25.03%)263.00 (+16.37%)166.00 (+4.53%)144.00 (+1.41%)126.00 (-5.26%)49.38 (+43.17%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)226.00 (n/a)158.80 (n/a)142.00 (n/a)133.00 (n/a)34.49 (n/a)

mem_copy_1_cores_1_chans_2048_tile_2048_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.06 (-29.56%)0.05 (-22.18%)0.06 (-11.77%)0.04 (-28.64%)0.01 (-18.98%)210.50 (+40.15%)156.62 (+28.71%)142.30 (+13.30%)139.00 (+41.98%)30.61 (+64.93%)
130b6ea — 2025-12-05 21:15:490.08 (n/a)0.07 (n/a)0.07 (n/a)0.05 (n/a)0.01 (n/a)150.20 (n/a)121.68 (n/a)125.60 (n/a)97.90 (n/a)18.56 (n/a)

mem_copy_1_cores_1_chans_2048_tile_2048_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (-0.24%)0.08 (-17.53%)0.08 (-27.05%)0.06 (-15.43%)0.02 (+4.62%)148.50 (+18.23%)113.02 (+23.38%)103.60 (+37.04%)74.10 (+0.14%)31.69 (+33.33%)
84d3478 — 2026-02-17 23:25:130.11 (n/a)0.09 (n/a)0.11 (n/a)0.07 (n/a)0.02 (n/a)125.60 (n/a)91.60 (n/a)75.60 (n/a)74.00 (n/a)23.77 (n/a)

mem_copy_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (-15.32%)0.05 (-17.09%)0.05 (-23.81%)0.04 (-15.54%)0.01 (-18.18%)193.00 (+18.40%)160.80 (+20.36%)168.00 (+31.25%)124.00 (+18.10%)22.92 (+11.39%)
0999a9a — 2025-11-13 19:24:080.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)163.00 (n/a)133.60 (n/a)128.00 (n/a)105.00 (n/a)20.58 (n/a)

mem_copy_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (-23.02%)0.05 (-6.98%)0.05 (-2.13%)0.04 (-15.31%)0.01 (-33.29%)196.00 (+18.07%)147.00 (+6.52%)141.00 (+2.17%)126.00 (+29.90%)25.27 (+8.13%)
0999a9a — 2025-11-13 19:24:080.08 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)166.00 (n/a)138.00 (n/a)138.00 (n/a)97.00 (n/a)23.37 (n/a)

mem_copy_2_cores_1_chans_2048_tile_1024_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.07 (-28.39%)0.05 (-38.49%)0.05 (-43.76%)0.04 (-46.17%)0.01 (+34.99%)204.10 (+85.71%)165.42 (+68.28%)179.30 (+77.88%)110.90 (+39.67%)36.90 (+250.07%)
130b6ea — 2025-12-05 21:15:490.10 (n/a)0.08 (n/a)0.08 (n/a)0.07 (n/a)0.01 (n/a)109.90 (n/a)98.30 (n/a)100.80 (n/a)79.40 (n/a)10.54 (n/a)

mem_copy_2_cores_1_chans_2048_tile_1024_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (+27.03%)0.09 (+25.79%)0.11 (+27.09%)0.05 (+12.74%)0.03 (+43.93%)161.90 (-11.29%)94.34 (-18.35%)76.30 (-21.34%)74.00 (-21.36%)38.01 (+0.49%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)182.50 (n/a)115.54 (n/a)97.00 (n/a)94.10 (n/a)37.82 (n/a)

mem_copy_2_cores_2_chans_2048_tile_1024_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.07 (-31.29%)0.04 (-43.83%)0.04 (-50.67%)0.03 (-40.94%)0.01 (+5.22%)254.20 (+69.35%)212.06 (+85.04%)226.50 (+102.78%)122.10 (+45.53%)53.83 (+153.49%)
130b6ea — 2025-12-05 21:15:490.10 (n/a)0.07 (n/a)0.07 (n/a)0.05 (n/a)0.01 (n/a)150.10 (n/a)114.60 (n/a)111.70 (n/a)83.90 (n/a)21.23 (n/a)

mem_copy_2_cores_2_chans_2048_tile_1024_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.10 (+2.70%)0.08 (-5.38%)0.09 (+2.10%)0.04 (-43.30%)0.02 (+170.50%)201.60 (+76.38%)112.08 (+16.53%)90.80 (-2.05%)84.60 (-2.65%)50.14 (+375.75%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.09 (n/a)0.09 (n/a)0.07 (n/a)0.01 (n/a)114.30 (n/a)96.18 (n/a)92.70 (n/a)86.90 (n/a)10.54 (n/a)

mem_copy_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.07 (+11.97%)0.05 (+3.09%)0.05 (+2.10%)0.03 (-11.35%)0.01 (+41.65%)229.00 (+12.81%)156.00 (-0.51%)143.00 (-2.05%)117.00 (-10.69%)38.21 (+48.71%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)203.00 (n/a)156.80 (n/a)146.00 (n/a)131.00 (n/a)25.69 (n/a)

mem_copy_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (+9.02%)0.05 (+5.59%)0.05 (+0.00%)0.04 (+3.63%)0.01 (+25.48%)193.00 (-3.50%)147.40 (-4.66%)147.00 (+0.00%)122.00 (-8.27%)25.24 (+7.48%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)200.00 (n/a)154.60 (n/a)147.00 (n/a)133.00 (n/a)23.48 (n/a)

mem_copy_4_cores_1_chans_2048_tile_512_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.06 (-32.43%)0.05 (-36.02%)0.05 (-34.74%)0.04 (-43.96%)0.01 (-8.34%)231.50 (+78.49%)179.50 (+58.43%)179.10 (+53.21%)133.80 (+48.01%)35.96 (+139.13%)
130b6ea — 2025-12-05 21:15:490.09 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)129.70 (n/a)113.30 (n/a)116.90 (n/a)90.40 (n/a)15.04 (n/a)

mem_copy_4_cores_1_chans_2048_tile_512_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.09 (-5.87%)0.08 (-3.61%)0.08 (-0.79%)0.06 (-12.74%)0.01 (+22.04%)146.30 (+14.66%)108.02 (+4.96%)99.30 (+0.81%)95.10 (+6.26%)21.49 (+48.38%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)127.60 (n/a)102.92 (n/a)98.50 (n/a)89.50 (n/a)14.49 (n/a)

mem_copy_4_cores_2_chans_2048_tile_512_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.07 (-17.65%)0.05 (-33.07%)0.04 (-41.73%)0.04 (-36.76%)0.01 (+31.61%)212.70 (+58.14%)179.54 (+53.93%)200.50 (+71.66%)119.00 (+21.43%)39.76 (+153.06%)
130b6ea — 2025-12-05 21:15:490.08 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)134.50 (n/a)116.64 (n/a)116.80 (n/a)98.00 (n/a)15.71 (n/a)

mem_copy_4_cores_2_chans_2048_tile_512_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (+13.04%)0.09 (+17.19%)0.10 (+20.43%)0.06 (+17.23%)0.02 (+16.34%)142.80 (-14.70%)91.94 (-14.71%)79.00 (-17.02%)77.70 (-11.50%)28.50 (-14.82%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)167.40 (n/a)107.80 (n/a)95.20 (n/a)87.80 (n/a)33.46 (n/a)

mem_copy_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (-13.48%)0.05 (-17.79%)0.05 (-15.33%)0.03 (-25.09%)0.01 (+6.67%)267.00 (+33.50%)176.80 (+24.51%)150.00 (+18.11%)141.00 (+15.57%)47.41 (+61.41%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.06 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)200.00 (n/a)142.00 (n/a)127.00 (n/a)122.00 (n/a)29.37 (n/a)

mem_copy_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (+13.38%)0.05 (+23.85%)0.05 (+37.91%)0.05 (+30.72%)0.00 (-56.85%)166.00 (-23.50%)153.40 (-20.60%)153.00 (-27.49%)142.00 (-11.80%)7.63 (-70.60%)
0999a9a — 2025-11-13 19:24:080.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)217.00 (n/a)193.20 (n/a)211.00 (n/a)161.00 (n/a)25.96 (n/a)

mem_copy_8_cores_1_chans_2048_tile_256_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.06 (-27.87%)0.04 (-41.92%)0.04 (-44.07%)0.03 (-56.96%)0.01 (+130.11%)272.50 (+132.31%)200.18 (+85.70%)194.10 (+78.89%)130.40 (+38.58%)62.46 (+653.01%)
130b6ea — 2025-12-05 21:15:490.09 (n/a)0.08 (n/a)0.08 (n/a)0.07 (n/a)0.01 (n/a)117.30 (n/a)107.80 (n/a)108.50 (n/a)94.10 (n/a)8.29 (n/a)

mem_copy_8_cores_1_chans_2048_tile_256_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-0.39%)0.07 (+10.07%)0.08 (+1.07%)0.06 (+38.38%)0.01 (-47.35%)136.70 (-27.75%)110.64 (-13.07%)104.70 (-1.13%)103.00 (+0.39%)14.59 (-60.78%)
84d3478 — 2026-02-17 23:25:130.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)189.20 (n/a)127.28 (n/a)105.90 (n/a)102.60 (n/a)37.20 (n/a)

mem_copy_8_cores_2_chans_2048_tile_256_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.05 (-47.39%)0.04 (-42.95%)0.04 (-39.47%)0.03 (-50.88%)0.01 (-36.21%)278.00 (+103.66%)213.08 (+76.30%)206.70 (+65.23%)174.00 (+89.96%)40.35 (+156.66%)
130b6ea — 2025-12-05 21:15:490.09 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)136.50 (n/a)120.86 (n/a)125.10 (n/a)91.60 (n/a)15.72 (n/a)

mem_copy_8_cores_2_chans_2048_tile_256_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-17.85%)0.07 (-24.66%)0.08 (-18.40%)0.04 (-32.08%)0.02 (+21.78%)185.30 (+47.18%)127.72 (+37.87%)102.00 (+22.45%)100.40 (+21.70%)38.66 (+106.34%)
84d3478 — 2026-02-17 23:25:130.10 (n/a)0.09 (n/a)0.10 (n/a)0.07 (n/a)0.01 (n/a)125.90 (n/a)92.64 (n/a)83.30 (n/a)82.50 (n/a)18.74 (n/a)

mha

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.18 (-0.57%)0.18 (-0.61%)0.18 (-0.56%)0.18 (-0.76%)0.00 (+154.51%)47616.50 (+0.77%)47504.24 (+0.62%)47478.70 (+0.57%)47436.00 (+0.57%)74.08 (+158.08%)
130b6ea — 2025-12-05 21:15:490.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47253.90 (n/a)47213.70 (n/a)47210.90 (n/a)47167.70 (n/a)28.71 (n/a)

mha0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
84d3478 — 2026-02-17 23:25:130.18 (-0.51%)0.18 (-0.36%)0.18 (-0.41%)0.18 (-0.21%)0.00 (-35.21%)47468.60 (+0.21%)47403.94 (+0.36%)47426.00 (+0.41%)47314.40 (+0.51%)68.83 (-34.73%)
1531781 — 2026-01-23 18:49:240.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47367.50 (n/a)47232.68 (n/a)47232.00 (n/a)47073.00 (n/a)105.45 (n/a)

mha_16384_64_1_8_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47533.70 (n/a)47380.56 (n/a)47337.90 (n/a)47268.00 (n/a)110.60 (n/a)

relu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.06 (-27.04%)0.05 (-34.23%)0.05 (-46.94%)0.04 (-27.93%)0.01 (-47.94%)209.60 (+38.81%)171.00 (+47.87%)176.00 (+88.44%)126.10 (+37.07%)30.47 (-3.44%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.07 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)151.00 (n/a)115.64 (n/a)93.40 (n/a)92.00 (n/a)31.56 (n/a)

relu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.07 (-25.04%)0.05 (-35.32%)0.06 (-36.66%)0.04 (-43.89%)0.01 (+10.24%)232.90 (+78.19%)165.38 (+59.57%)148.80 (+57.96%)124.50 (+33.44%)42.96 (+167.53%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.09 (n/a)0.06 (n/a)0.01 (n/a)130.70 (n/a)103.64 (n/a)94.20 (n/a)93.30 (n/a)16.06 (n/a)

relu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.06 (-48.94%)0.04 (-52.23%)0.05 (-53.98%)0.03 (-38.40%)0.01 (-61.44%)280.40 (+62.36%)193.56 (+98.73%)170.20 (+117.37%)148.10 (+95.90%)51.99 (+23.25%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.10 (n/a)0.05 (n/a)0.03 (n/a)172.70 (n/a)97.40 (n/a)78.30 (n/a)75.60 (n/a)42.18 (n/a)

relu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.06 (-23.14%)0.05 (-32.20%)0.05 (-38.79%)0.04 (-12.52%)0.01 (-44.78%)231.50 (+14.32%)174.38 (+41.73%)168.80 (+63.41%)131.80 (+30.11%)36.46 (-17.96%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)202.50 (n/a)123.04 (n/a)103.30 (n/a)101.30 (n/a)44.44 (n/a)

rms_norm_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.06 (-34.42%)0.05 (-37.04%)0.05 (-44.22%)0.04 (-21.86%)0.00 (-63.65%)182.90 (+27.99%)170.52 (+56.10%)174.50 (+79.34%)146.60 (+52.39%)14.23 (-29.30%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)142.90 (n/a)109.24 (n/a)97.30 (n/a)96.20 (n/a)20.13 (n/a)

rms_norm_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.06 (-30.65%)0.05 (-36.30%)0.05 (-44.98%)0.04 (-28.49%)0.01 (-30.71%)205.30 (+39.85%)169.38 (+56.66%)178.60 (+81.87%)139.70 (+44.17%)28.78 (+32.63%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)146.80 (n/a)108.12 (n/a)98.20 (n/a)96.90 (n/a)21.70 (n/a)

rms_norm_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.06 (-50.33%)0.05 (-48.25%)0.05 (-49.18%)0.04 (-38.82%)0.01 (-63.04%)204.60 (+63.42%)160.88 (+88.52%)153.50 (+96.79%)137.10 (+101.32%)26.49 (+16.78%)
897d04e — 2026-03-06 22:12:250.12 (n/a)0.10 (n/a)0.10 (n/a)0.07 (n/a)0.02 (n/a)125.20 (n/a)85.34 (n/a)78.00 (n/a)68.10 (n/a)22.68 (n/a)

rms_norm_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.06 (-29.81%)0.05 (-41.44%)0.05 (-44.67%)0.03 (-44.32%)0.01 (-3.45%)251.50 (+79.64%)186.06 (+75.53%)176.20 (+80.72%)137.80 (+42.50%)46.59 (+145.00%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)140.00 (n/a)106.00 (n/a)97.50 (n/a)96.70 (n/a)19.02 (n/a)

rms_norm_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.07 (-22.17%)0.05 (-36.23%)0.05 (-44.38%)0.04 (-34.61%)0.01 (-16.06%)232.70 (+52.89%)168.60 (+58.67%)171.90 (+79.81%)119.90 (+28.51%)41.98 (+63.32%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.01 (n/a)152.20 (n/a)106.26 (n/a)95.60 (n/a)93.30 (n/a)25.71 (n/a)

rms_norm_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.05 (-39.11%)0.04 (-47.72%)0.04 (-53.20%)0.04 (-43.83%)0.01 (-28.59%)232.20 (+78.07%)199.26 (+92.56%)208.40 (+113.74%)157.10 (+64.16%)31.18 (+106.75%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)130.40 (n/a)103.48 (n/a)97.50 (n/a)95.70 (n/a)15.08 (n/a)

rms_norm_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.05 (-40.83%)0.04 (-42.13%)0.04 (-48.58%)0.04 (-10.61%)0.00 (-70.79%)212.90 (+11.88%)195.54 (+63.71%)201.90 (+94.51%)160.90 (+69.01%)20.56 (-48.33%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)190.30 (n/a)119.44 (n/a)103.80 (n/a)95.20 (n/a)39.79 (n/a)

rms_norm_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.05 (-56.54%)0.04 (-58.45%)0.04 (-63.82%)0.03 (-27.43%)0.01 (-78.46%)271.30 (+37.79%)227.02 (+116.42%)229.60 (+176.29%)179.10 (+130.21%)33.05 (-35.81%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.10 (n/a)0.04 (n/a)0.03 (n/a)196.90 (n/a)104.90 (n/a)83.10 (n/a)77.80 (n/a)51.49 (n/a)

rope_1_cols_2_channels_4096_tile_4096_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.19 (-21.66%)0.17 (-8.53%)0.18 (+2.09%)0.14 (-6.71%)0.02 (-38.07%)174.10 (+7.20%)150.06 (+8.11%)139.60 (-2.04%)129.80 (+27.63%)20.36 (-10.65%)
d48746f — 2026-01-12 22:17:100.24 (n/a)0.18 (n/a)0.17 (n/a)0.15 (n/a)0.04 (n/a)162.40 (n/a)138.80 (n/a)142.50 (n/a)101.70 (n/a)22.79 (n/a)

rope_1c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.88 (-17.19%)0.67 (-24.63%)0.71 (-21.72%)0.42 (-41.23%)0.18 (+24.11%)235.30 (+70.14%)157.78 (+38.96%)138.00 (+27.66%)111.80 (+20.73%)49.36 (+157.34%)
897d04e — 2026-03-06 22:12:251.06 (n/a)0.89 (n/a)0.91 (n/a)0.71 (n/a)0.15 (n/a)138.30 (n/a)113.54 (n/a)108.10 (n/a)92.60 (n/a)19.18 (n/a)

rope_1c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.62 (+1.19%)0.47 (-6.74%)0.40 (-17.58%)0.35 (-20.83%)0.12 (+82.73%)209.30 (+26.31%)165.28 (+11.36%)184.80 (+21.34%)118.30 (-1.17%)39.40 (+130.37%)
897d04e — 2026-03-06 22:12:250.62 (n/a)0.50 (n/a)0.48 (n/a)0.44 (n/a)0.07 (n/a)165.70 (n/a)148.42 (n/a)152.30 (n/a)119.70 (n/a)17.10 (n/a)

rope_2_cols_2_channels_4096_tile_2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.21 (-0.56%)0.17 (-10.62%)0.16 (-13.58%)0.13 (-13.33%)0.03 (+17.05%)183.70 (+15.39%)151.76 (+12.77%)155.20 (+15.73%)117.30 (+0.60%)24.24 (+35.73%)
d48746f — 2026-01-12 22:17:100.21 (n/a)0.19 (n/a)0.18 (n/a)0.15 (n/a)0.02 (n/a)159.20 (n/a)134.58 (n/a)134.10 (n/a)116.60 (n/a)17.86 (n/a)

rope_2c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.83 (-22.91%)0.64 (-28.95%)0.63 (-31.74%)0.47 (-16.75%)0.14 (-28.30%)207.20 (+20.12%)160.98 (+39.21%)155.20 (+46.55%)118.00 (+29.67%)36.35 (+9.85%)
897d04e — 2026-03-06 22:12:251.08 (n/a)0.90 (n/a)0.93 (n/a)0.57 (n/a)0.20 (n/a)172.50 (n/a)115.64 (n/a)105.90 (n/a)91.00 (n/a)33.09 (n/a)

rope_2c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.58 (-33.23%)0.47 (-25.11%)0.45 (-11.52%)0.38 (-21.84%)0.08 (-56.07%)193.80 (+27.92%)159.74 (+28.76%)162.40 (+13.01%)126.10 (+49.76%)25.53 (-17.56%)
897d04e — 2026-03-06 22:12:250.88 (n/a)0.63 (n/a)0.51 (n/a)0.49 (n/a)0.18 (n/a)151.50 (n/a)124.06 (n/a)143.70 (n/a)84.20 (n/a)30.96 (n/a)

rope_4_cols_2_channels_4096_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.22 (+18.92%)0.17 (+6.24%)0.17 (+5.57%)0.14 (-0.50%)0.03 (+90.77%)171.10 (+0.47%)147.40 (-4.39%)148.10 (-5.31%)110.10 (-15.89%)23.59 (+60.67%)
d48746f — 2026-01-12 22:17:100.19 (n/a)0.16 (n/a)0.16 (n/a)0.14 (n/a)0.02 (n/a)170.30 (n/a)154.16 (n/a)156.40 (n/a)130.90 (n/a)14.69 (n/a)

rope_8_cols_2_channels_4096_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.17 (-1.30%)0.15 (+6.63%)0.15 (-8.78%)0.14 (+39.68%)0.01 (-67.35%)174.90 (-28.41%)163.42 (-10.36%)167.40 (+9.63%)146.50 (+1.31%)11.19 (-76.17%)
d48746f — 2026-01-12 22:17:100.17 (n/a)0.14 (n/a)0.16 (n/a)0.10 (n/a)0.03 (n/a)244.30 (n/a)182.30 (n/a)152.70 (n/a)144.60 (n/a)46.98 (n/a)

rope_8c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.75 (-29.39%)0.53 (-39.73%)0.52 (-49.22%)0.28 (-48.50%)0.17 (-24.84%)355.00 (+94.20%)206.04 (+72.88%)188.80 (+96.87%)130.70 (+41.60%)86.96 (+124.18%)
897d04e — 2026-03-06 22:12:251.07 (n/a)0.88 (n/a)1.03 (n/a)0.54 (n/a)0.23 (n/a)182.80 (n/a)119.18 (n/a)95.90 (n/a)92.30 (n/a)38.79 (n/a)

rope_8c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.54 (-3.69%)0.40 (-16.59%)0.37 (-29.12%)0.27 (-26.23%)0.10 (+16.97%)275.10 (+35.52%)194.68 (+22.98%)198.10 (+41.10%)136.40 (+3.88%)52.10 (+66.05%)
897d04e — 2026-03-06 22:12:250.56 (n/a)0.48 (n/a)0.52 (n/a)0.36 (n/a)0.09 (n/a)203.00 (n/a)158.30 (n/a)140.40 (n/a)131.30 (n/a)31.38 (n/a)

sigmoid_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.06 (-35.79%)0.05 (-36.61%)0.05 (-44.47%)0.04 (-23.34%)0.01 (-64.95%)197.40 (+30.47%)174.10 (+52.69%)169.50 (+80.13%)145.70 (+55.83%)20.72 (-26.10%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)151.30 (n/a)114.02 (n/a)94.10 (n/a)93.50 (n/a)28.04 (n/a)

sigmoid_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.06 (-47.28%)0.04 (-50.17%)0.04 (-57.93%)0.03 (-30.15%)0.01 (-65.09%)252.70 (+43.17%)199.38 (+86.69%)182.90 (+137.84%)145.00 (+89.79%)47.05 (+5.00%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.11 (n/a)0.05 (n/a)0.03 (n/a)176.50 (n/a)106.80 (n/a)76.90 (n/a)76.40 (n/a)44.81 (n/a)

sigmoid_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.06 (-36.12%)0.05 (-38.90%)0.04 (-49.66%)0.04 (-17.17%)0.01 (-46.78%)207.20 (+20.75%)180.26 (+59.64%)199.10 (+98.70%)141.40 (+56.59%)32.12 (-3.50%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.08 (n/a)0.05 (n/a)0.02 (n/a)171.60 (n/a)112.92 (n/a)100.20 (n/a)90.30 (n/a)33.29 (n/a)

sigmoid_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.05 (-46.43%)0.04 (-51.06%)0.04 (-58.68%)0.03 (-11.06%)0.01 (-73.52%)239.60 (+12.44%)197.88 (+81.24%)201.90 (+142.09%)154.30 (+86.80%)31.65 (-45.52%)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.09 (n/a)0.10 (n/a)0.04 (n/a)0.03 (n/a)213.10 (n/a)109.18 (n/a)83.40 (n/a)82.60 (n/a)58.09 (n/a)

silu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.07 (-38.79%)0.05 (-38.81%)0.05 (-53.01%)0.04 (-1.72%)0.01 (-70.28%)191.90 (+1.75%)163.82 (+45.23%)167.40 (+112.98%)123.70 (+63.41%)26.02 (-49.01%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.08 (n/a)0.10 (n/a)0.04 (n/a)0.03 (n/a)188.60 (n/a)112.80 (n/a)78.60 (n/a)75.70 (n/a)51.04 (n/a)

silu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.06 (-32.61%)0.05 (-29.45%)0.05 (-31.48%)0.04 (-20.94%)0.01 (-55.59%)182.60 (+26.45%)155.16 (+38.56%)158.80 (+45.96%)129.20 (+48.34%)20.16 (-16.21%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.02 (n/a)144.40 (n/a)111.98 (n/a)108.80 (n/a)87.10 (n/a)24.06 (n/a)

silu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.07 (-22.51%)0.04 (-42.27%)0.04 (-49.53%)0.03 (-44.68%)0.02 (-2.23%)299.90 (+80.77%)200.28 (+80.99%)196.20 (+98.18%)115.90 (+29.06%)65.90 (+111.55%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.08 (n/a)0.05 (n/a)0.02 (n/a)165.90 (n/a)110.66 (n/a)99.00 (n/a)89.80 (n/a)31.15 (n/a)

silu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.05 (-49.46%)0.05 (-44.30%)0.05 (-47.96%)0.04 (-5.09%)0.00 (-85.49%)192.60 (+5.36%)172.24 (+64.16%)167.60 (+92.20%)161.20 (+97.79%)12.73 (-70.83%)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.09 (n/a)0.09 (n/a)0.04 (n/a)0.02 (n/a)182.80 (n/a)104.92 (n/a)87.20 (n/a)81.50 (n/a)43.64 (n/a)

softmax_1_cols_2_channels_4096_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.19 (-8.94%)0.17 (-8.66%)0.17 (-19.83%)0.14 (+4.32%)0.02 (-35.13%)115.80 (-4.14%)99.26 (+7.47%)98.60 (+24.65%)85.00 (+9.82%)13.65 (-30.86%)
84d3478 — 2026-02-17 23:25:130.21 (n/a)0.18 (n/a)0.21 (n/a)0.14 (n/a)0.04 (n/a)120.80 (n/a)92.36 (n/a)79.10 (n/a)77.40 (n/a)19.75 (n/a)

softmax_2_cols_2_channels_32768_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:571.12 (n/a)0.73 (n/a)0.66 (n/a)0.56 (n/a)0.23 (n/a)234.40 (n/a)191.46 (n/a)197.60 (n/a)117.10 (n/a)47.84 (n/a)

softmax_2_cols_2_channels_32768_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:571.28 (n/a)0.77 (n/a)0.64 (n/a)0.57 (n/a)0.30 (n/a)230.80 (n/a)186.26 (n/a)205.80 (n/a)102.30 (n/a)54.79 (n/a)

softmax_2_cols_2_channels_4096_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.20 (-0.87%)0.16 (+6.57%)0.17 (-9.55%)0.11 (+29.81%)0.03 (-44.34%)145.70 (-22.99%)104.68 (-15.00%)95.00 (+10.59%)83.80 (+0.84%)24.16 (-54.23%)
84d3478 — 2026-02-17 23:25:130.20 (n/a)0.15 (n/a)0.19 (n/a)0.09 (n/a)0.06 (n/a)189.20 (n/a)123.16 (n/a)85.90 (n/a)83.10 (n/a)52.80 (n/a)

softmax_2_cols_2_channels_4096_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.25 (+23.53%)0.18 (+9.42%)0.18 (-3.86%)0.11 (+2.99%)0.05 (+19.24%)144.10 (-2.90%)98.98 (-8.18%)92.10 (+4.07%)64.30 (-19.02%)29.58 (-7.13%)
84d3478 — 2026-02-17 23:25:130.21 (n/a)0.16 (n/a)0.19 (n/a)0.11 (n/a)0.04 (n/a)148.40 (n/a)107.80 (n/a)88.50 (n/a)79.40 (n/a)31.85 (n/a)

softmax_4_cols_4_channels_32768_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.82 (n/a)0.60 (n/a)0.53 (n/a)0.52 (n/a)0.13 (n/a)251.30 (n/a)224.02 (n/a)246.00 (n/a)160.50 (n/a)38.87 (n/a)

swiglu

No metrics available.

swiglu_decode_1x2048x2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:57n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)5187.20 (n/a)4720.84 (n/a)4750.90 (n/a)4015.10 (n/a)390.23 (n/a)

swiglu_decode_1x2048x2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.00 (-0.21%)0.00 (+7.74%)0.00 (+0.37%)0.00 (+8.84%)0.00 (-19.58%)5170.80 (-8.12%)3761.80 (-8.70%)3408.60 (-0.36%)3407.30 (+0.21%)787.66 (-22.98%)
84d3478 — 2026-02-17 23:25:130.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)5627.70 (n/a)4120.44 (n/a)3421.00 (n/a)3400.30 (n/a)1022.61 (n/a)

swiglu_prefill_256x2048x2048

No metrics available.

tanh_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.07 (-18.40%)0.05 (-34.25%)0.04 (-42.63%)0.04 (-37.24%)0.01 (+40.12%)206.70 (+59.37%)167.64 (+57.38%)183.10 (+74.38%)116.80 (+22.56%)38.05 (+172.81%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)129.70 (n/a)106.52 (n/a)105.00 (n/a)95.30 (n/a)13.95 (n/a)

tanh_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.06 (-45.24%)0.04 (-52.25%)0.04 (-62.52%)0.04 (-30.20%)0.01 (-56.72%)229.10 (+43.28%)190.58 (+100.78%)203.10 (+166.89%)136.40 (+82.60%)41.66 (+13.42%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.11 (n/a)0.05 (n/a)0.02 (n/a)159.90 (n/a)94.92 (n/a)76.10 (n/a)74.70 (n/a)36.73 (n/a)

tanh_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.06 (-24.14%)0.05 (-31.11%)0.05 (-38.07%)0.04 (-36.28%)0.01 (+4.46%)202.40 (+56.90%)165.44 (+46.93%)172.10 (+61.44%)133.50 (+31.79%)27.67 (+113.37%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.07 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)129.00 (n/a)112.60 (n/a)106.60 (n/a)101.30 (n/a)12.97 (n/a)

tanh_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.05 (-40.73%)0.04 (-44.47%)0.04 (-49.40%)0.04 (-40.85%)0.00 (-47.53%)226.00 (+69.04%)197.56 (+79.60%)201.50 (+97.74%)167.60 (+68.78%)21.51 (+48.49%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)133.70 (n/a)110.00 (n/a)101.90 (n/a)99.30 (n/a)14.49 (n/a)

transpose_2048_M_64_N_1_cols_1_channels_64_m_64_n_8_s

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:573.08 (-21.45%)2.36 (-35.84%)2.28 (-37.62%)1.87 (-45.86%)0.48 (+215.50%)280.40 (+84.72%)229.08 (+60.53%)229.80 (+60.25%)170.00 (+27.34%)43.70 (+641.06%)
130b6ea — 2025-12-05 21:15:493.93 (n/a)3.68 (n/a)3.66 (n/a)3.45 (n/a)0.15 (n/a)151.80 (n/a)142.70 (n/a)143.40 (n/a)133.50 (n/a)5.90 (n/a)

transpose_2048_M_64_N_1_cols_1_channels_64_m_64_n_8_s0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:254.43 (+0.85%)3.73 (+3.83%)4.02 (-1.30%)2.97 (+16.77%)0.65 (-24.99%)176.70 (-14.35%)144.46 (-6.24%)130.40 (+1.32%)118.40 (-0.84%)26.60 (-35.16%)
84d3478 — 2026-02-17 23:25:134.39 (n/a)3.59 (n/a)4.07 (n/a)2.54 (n/a)0.87 (n/a)206.30 (n/a)154.08 (n/a)128.70 (n/a)119.40 (n/a)41.02 (n/a)

transpose_2048_M_64_N_1_cols_2_channels_64_m_64_n_8_s

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:573.02 (-40.79%)2.61 (-32.22%)2.57 (-24.37%)2.23 (-32.11%)0.33 (-53.32%)235.10 (+47.31%)203.28 (+45.01%)203.80 (+32.17%)173.30 (+68.91%)25.75 (+13.43%)
130b6ea — 2025-12-05 21:15:495.11 (n/a)3.85 (n/a)3.40 (n/a)3.29 (n/a)0.71 (n/a)159.60 (n/a)140.18 (n/a)154.20 (n/a)102.60 (n/a)22.70 (n/a)

transpose_2048_M_64_N_1_cols_2_channels_64_m_64_n_8_s0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:255.27 (-14.18%)4.35 (-2.49%)4.45 (+2.15%)2.78 (-22.18%)0.98 (-3.38%)188.60 (+28.47%)126.98 (+4.20%)117.90 (-2.08%)99.40 (+16.53%)36.02 (+50.15%)
84d3478 — 2026-02-17 23:25:136.14 (n/a)4.46 (n/a)4.35 (n/a)3.57 (n/a)1.02 (n/a)146.80 (n/a)121.86 (n/a)120.40 (n/a)85.30 (n/a)23.99 (n/a)

weighted_rms_norm_1_cols_2_channels_2048_weights_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.11 (+30.60%)0.07 (-0.75%)0.07 (-5.65%)0.05 (-6.47%)0.02 (+70.41%)271.60 (+60.43%)191.50 (+58.03%)179.50 (+58.99%)114.70 (+14.81%)58.30 (+104.50%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.07 (n/a)0.07 (n/a)0.05 (n/a)0.01 (n/a)169.30 (n/a)121.18 (n/a)112.90 (n/a)99.90 (n/a)28.51 (n/a)

weighted_rms_norm_2_cols_2_channels_2048_weights_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.06 (-44.63%)0.05 (-43.78%)0.05 (-47.53%)0.05 (-16.19%)0.00 (-82.32%)209.10 (+49.14%)191.98 (+111.43%)187.60 (+138.07%)176.00 (+125.93%)13.12 (-52.50%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.10 (n/a)0.10 (n/a)0.06 (n/a)0.02 (n/a)140.20 (n/a)90.80 (n/a)78.80 (n/a)77.90 (n/a)27.62 (n/a)

weighted_rms_norm_4_cols_2_channels_2048_weights_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.07 (-37.40%)0.04 (-55.41%)0.04 (-63.81%)0.02 (-67.84%)0.02 (+3.70%)443.20 (+249.80%)254.86 (+185.78%)247.70 (+210.79%)138.80 (+79.79%)118.78 (+463.05%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.10 (n/a)0.10 (n/a)0.06 (n/a)0.02 (n/a)126.70 (n/a)89.18 (n/a)79.70 (n/a)77.20 (n/a)21.10 (n/a)

weighted_rms_norm_8_cols_2_channels_2048_weights_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
27cd2f3 — 2026-03-20 17:09:570.05 (-43.26%)0.04 (-43.05%)0.04 (-47.23%)0.04 (-21.17%)0.00 (-74.72%)239.60 (+34.76%)214.32 (+79.47%)209.30 (+101.25%)192.60 (+87.17%)19.50 (-40.41%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.07 (n/a)0.08 (n/a)0.05 (n/a)0.01 (n/a)177.80 (n/a)119.42 (n/a)104.00 (n/a)102.90 (n/a)32.72 (n/a)

@andrej andrej marked this pull request as ready for review March 20, 2026 17:21
@github-actions
Copy link
Contributor

📊 Test Results for Test Example Applications

27cd2f3 (2026_03_20_17_21_26)

IRONCLAD

Tested on 2026_03_20_17_21_26 at commit 27cd2f3.

Test Checks TTFT (mean)TPS (mean)
llama_3.2_1b_prompt_1024_tokens_40 ✅ 1/1 2.34 n/a
📈 Trends (vs main branch) for Test Example Applications

27cd2f3 (2026_03_20_17_21_26)

IRONCLAD Trends

llama_3.2_1b

Commit/Date Num Tokens (max)Num Tokens (mean)Num Tokens (median)Num Tokens (min)Num Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)Total (max)Total (mean)Total (median)Total (min)Total (stddev)
130b6ea — 2025-12-05 21:33:1240.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.71 (-0.42%)4.64 (-0.09%)4.64 (+0.65%)4.55 (-0.22%)0.05 (-17.66%)4.41 (-0.34%)4.39 (-0.19%)4.38 (-0.33%)4.37 (-0.15%)0.01 (-25.90%)12.96 (-0.00%)12.80 (+0.07%)12.80 (-0.23%)12.67 (+0.44%)0.09 (-21.12%)
0a6c11c — 2025-12-03 23:35:1540.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.73 (n/a)4.64 (n/a)4.61 (n/a)4.56 (n/a)0.06 (n/a)4.42 (n/a)4.40 (n/a)4.40 (n/a)4.37 (n/a)0.02 (n/a)12.96 (n/a)12.79 (n/a)12.83 (n/a)12.62 (n/a)0.12 (n/a)

llama_3.2_1b_prompt_1024_tokens_40

Commit/Date TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
27cd2f3 — 2026-03-20 17:19:252.34 (n/a)2.34 (n/a)2.34 (n/a)2.34 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_13_tokens_1

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:071.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.62 (-4.45%)0.62 (-4.45%)0.62 (-4.45%)0.62 (-4.45%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:231.00 (n/a)1.00 (n/a)1.00 (n/a)1.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.65 (n/a)0.65 (n/a)0.65 (n/a)0.65 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_13_tokens_40

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:0740.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.30 (+2.38%)4.30 (+2.38%)4.30 (+2.38%)4.30 (+2.38%)0.00 (n/a)0.61 (-2.49%)0.61 (-2.49%)0.61 (-2.49%)0.61 (-2.49%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:2340.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.20 (n/a)4.20 (n/a)4.20 (n/a)4.20 (n/a)0.00 (n/a)0.63 (n/a)0.63 (n/a)0.63 (n/a)0.63 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_2048_tokens_1

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:071.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:231.00 (n/a)1.00 (n/a)1.00 (n/a)1.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_2048_tokens_40

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:0740.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)0.00 (n/a)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:2340.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)0.00 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)0.00 (n/a)

Copy link
Collaborator

@hunhoffe hunhoffe left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Look good! Thanks for all the debug work to get this out!

aie-opt --version
echo "========================"

- name: Debug - Print resource limits and environment
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do we still want to keep all this info? It doesn't seem harmful, but wanted to ask.

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

i'll clean it up a bit at least


void mask_bf16(bfloat16 *inout, const int32 unmasked_size, const int32 total_size)
{
// TODO: Optimize this to use vector code
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Should we make an issue for this TODO?

prompt_lengths = [2048, 13]
num_tokens_list = [40, 1]
prompt_lengths = [
1024, # 13
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

What are the commented out numbers?

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

oops. these are other llama configurations that we should also test. good catch -- i commented those out during debugging

@github-actions
Copy link
Contributor

📊 Test Results for Test Example Applications

e73d4ff (2026_03_20_22_01_21)

IRONCLAD

Tested on 2026_03_20_22_01_21 at commit e73d4ff.

Test Checks TTFT (mean)TPS (mean)
llama_3.2_1b_prompt_1024_tokens_1 ✅ 5/5 2.23 n/a
llama_3.2_1b_prompt_1024_tokens_40 ✅ 5/5 2.26 n/a
llama_3.2_1b_prompt_13_tokens_1 ✅ 5/5 2.16 n/a
llama_3.2_1b_prompt_13_tokens_40 ✅ 5/5 2.17 n/a
📈 Trends (vs main branch) for Test Example Applications

e73d4ff (2026_03_20_22_01_21)

IRONCLAD Trends

llama_3.2_1b

Commit/Date Num Tokens (max)Num Tokens (mean)Num Tokens (median)Num Tokens (min)Num Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)Total (max)Total (mean)Total (median)Total (min)Total (stddev)
130b6ea — 2025-12-05 21:33:1240.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.71 (-0.42%)4.64 (-0.09%)4.64 (+0.65%)4.55 (-0.22%)0.05 (-17.66%)4.41 (-0.34%)4.39 (-0.19%)4.38 (-0.33%)4.37 (-0.15%)0.01 (-25.90%)12.96 (-0.00%)12.80 (+0.07%)12.80 (-0.23%)12.67 (+0.44%)0.09 (-21.12%)
0a6c11c — 2025-12-03 23:35:1540.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.73 (n/a)4.64 (n/a)4.61 (n/a)4.56 (n/a)0.06 (n/a)4.42 (n/a)4.40 (n/a)4.40 (n/a)4.37 (n/a)0.02 (n/a)12.96 (n/a)12.79 (n/a)12.83 (n/a)12.62 (n/a)0.12 (n/a)

llama_3.2_1b_prompt_1024_tokens_1

Commit/Date TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
e73d4ff — 2026-03-20 21:55:502.24 (n/a)2.23 (n/a)2.23 (n/a)2.21 (n/a)0.01 (n/a)

llama_3.2_1b_prompt_1024_tokens_40

Commit/Date TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
e73d4ff — 2026-03-20 21:55:502.38 (n/a)2.26 (n/a)2.23 (n/a)2.22 (n/a)0.07 (n/a)

llama_3.2_1b_prompt_13_tokens_1

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
e73d4ff — 2026-03-20 21:55:50n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)2.18 (+251.64%)2.16 (+248.39%)2.16 (+248.58%)2.14 (+244.88%)0.02 (n/a)
897d04e — 2026-03-06 22:56:071.00 (n/a)1.00 (n/a)1.00 (n/a)1.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.62 (n/a)0.62 (n/a)0.62 (n/a)0.62 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_13_tokens_40

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
e73d4ff — 2026-03-20 21:55:50n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)2.18 (+255.11%)2.17 (+252.70%)2.17 (+253.15%)2.15 (+250.22%)0.01 (n/a)
897d04e — 2026-03-06 22:56:0740.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.30 (n/a)4.30 (n/a)4.30 (n/a)4.30 (n/a)0.00 (n/a)0.61 (n/a)0.61 (n/a)0.61 (n/a)0.61 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_2048_tokens_1

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:071.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:231.00 (n/a)1.00 (n/a)1.00 (n/a)1.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_2048_tokens_40

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:0740.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)0.00 (n/a)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:2340.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)0.00 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)0.00 (n/a)

@github-actions
Copy link
Contributor

📊 Test Results for Small Benchmark/Test Suite

e73d4ff (2026_03_20_22_10_55)

IRONCLAD

Tested on 2026_03_20_22_10_55 at commit e73d4ff.

Test Checks Latency (mean)Bandwidth (mean)GFLOP/s (mean)
axpy_1_cols_2_channels_2048_tile_2048_3.0 ✅ 5/5 175.30 0.07 n/a
axpy_2_cols_2_channels_2048_tile_1024_3.0 ✅ 5/5 208.80 0.06 n/a
axpy_4_cols_2_channels_2048_tile_512_3.0 ✅ 5/5 179.26 0.07 n/a
axpy_8_cols_2_channels_2048_tile_256_3.0 ✅ 5/5 198.44 0.06 n/a
dequant_1_cols_1_channels_2048_tile_2048 ✅ 5/5 181.14 0.03 n/a
dequant_1_cols_2_channels_2048_tile_1024 ✅ 5/5 160.98 0.03 n/a
dequant_2_cols_1_channels_2048_tile_1024 ✅ 5/5 151.02 0.04 n/a
dequant_2_cols_2_channels_2048_tile_512 ✅ 5/5 178.58 0.03 n/a
dequant_4_cols_1_channels_2048_tile_512 ✅ 5/5 193.62 0.03 n/a
dequant_4_cols_2_channels_2048_tile_256 ✅ 5/5 175.50 0.03 n/a
dequant_8_cols_1_channels_2048_tile_256 ✅ 5/5 201.66 0.03 n/a
dequant_8_cols_2_channels_2048_tile_128 ✅ 5/5 218.22 0.02 n/a
eltwise_add_1_cols_2_channels_2048_tile_2048 ✅ 5/5 185.30 0.07 n/a
eltwise_add_2_cols_2_channels_2048_tile_1024 ✅ 5/5 186.26 0.07 n/a
eltwise_add_4_cols_2_channels_2048_tile_512 ✅ 5/5 176.58 0.07 n/a
eltwise_add_8_cols_2_channels_2048_tile_256 ✅ 5/5 178.48 0.07 n/a
eltwise_mul_1_cols_2_channels_2048_tile_2048 ✅ 5/5 165.50 0.08 n/a
eltwise_mul_2_cols_2_channels_2048_tile_1024 ✅ 5/5 164.54 0.08 n/a
eltwise_mul_4_cols_2_channels_2048_tile_512 ✅ 5/5 263.78 0.05 n/a
eltwise_mul_8_cols_2_channels_2048_tile_256 ✅ 5/5 242.72 0.05 n/a
gelu_1_cols_1_channels_2048_tile_2048 ✅ 5/5 199.20 0.04 n/a
gelu_1_cols_2_channels_2048_tile_1024 ✅ 5/5 196.44 0.04 n/a
gelu_2_cols_1_channels_2048_tile_1024 ✅ 5/5 178.28 0.05 n/a
gelu_2_cols_2_channels_2048_tile_512 ✅ 5/5 176.86 0.05 n/a
gelu_4_cols_1_channels_2048_tile_512 ✅ 5/5 184.88 0.05 n/a
gelu_4_cols_2_channels_2048_tile_256 ✅ 5/5 182.48 0.05 n/a
gelu_8_cols_1_channels_2048_tile_256 ✅ 5/5 198.58 0.04 n/a
gelu_8_cols_2_channels_2048_tile_128 ✅ 5/5 252.14 0.03 n/a
gemm_1792x896x1152_64x32x48_8cols_ccolmaj ✅ 5/5 2162.06 4.38 n/a
gemm_192x384x64_48x96x16_4cols ✅ 5/5 276.14 0.88 n/a
gemm_192x384x64_48x96x16_4cols_bcolmaj_ccolmaj ✅ 5/5 285.48 0.84 n/a
gemm_2048x2048x2048_64x64x64_1cols ✅ 5/5 49398.32 0.51 n/a
gemm_2048x2048x2048_64x64x64_2cols_bcolmaj ✅ 5/5 28712.80 0.88 n/a
gemm_2048x2048x2048_64x64x64_8cols_bcolmaj_ccolmaj ✅ 5/5 7607.48 3.31 n/a
gemm_384x1536x1792_32x48x64_4cols_bcolmaj ✅ 5/5 2116.42 3.82 n/a
gemm_896x1792x640_32x64x80_8cols_ccolmaj ✅ 5/5 1326.72 5.12 n/a
layer_norm_1_cols_1_channels_2048_tile_2048 ✅ 5/5 180.20 0.05 n/a
layer_norm_1_cols_2_channels_2048_tile_1024 ✅ 5/5 176.44 0.05 n/a
layer_norm_2_cols_1_channels_2048_tile_1024 ✅ 5/5 148.68 0.06 n/a
layer_norm_2_cols_2_channels_2048_tile_512 ✅ 5/5 183.54 0.05 n/a
layer_norm_4_cols_1_channels_2048_tile_512 ✅ 5/5 172.04 0.05 n/a
layer_norm_4_cols_2_channels_2048_tile_256 ✅ 5/5 176.82 0.05 n/a
layer_norm_8_cols_1_channels_2048_tile_256 ✅ 5/5 164.82 0.05 n/a
layer_norm_8_cols_2_channels_2048_tile_128 ✅ 5/5 235.40 0.04 n/a
matrix_vector_mul_128x128_32tsi_128tso_1col ✅ 5/5 n/a 0.16 n/a
matrix_vector_mul_2048x8192_1tsi_1024tso_2col ✅ 5/5 n/a 23.81 n/a
matrix_vector_mul_2048x8192_1tsi_2048tso_1col ✅ 5/5 n/a 11.90 n/a
matrix_vector_mul_2048x8192_1tsi_256tso_8col ✅ 5/5 n/a 42.32 n/a
matrix_vector_mul_2048x8192_1tsi_512tso_4col ✅ 5/5 n/a 38.06 n/a
matrix_vector_mul_8192x2048_4tsi_1024tso_1col ✅ 5/5 n/a 12.79 n/a
matrix_vector_mul_8192x2048_4tsi_1024tso_2col ✅ 5/5 n/a 23.47 n/a
matrix_vector_mul_8192x2048_4tsi_1024tso_4col ✅ 5/5 n/a 39.63 n/a
matrix_vector_mul_8192x2048_4tsi_1024tso_8col ✅ 5/5 n/a 39.83 n/a
mem_copy_16_cores_2_chans_2048_tile_128_False ✅ 5/5 224.72 0.04 n/a
mem_copy_1_cores_1_chans_2048_tile_2048_False ✅ 5/5 150.86 0.06 n/a
mem_copy_2_cores_1_chans_2048_tile_1024_False ✅ 5/5 214.38 0.04 n/a
mem_copy_2_cores_2_chans_2048_tile_1024_False ✅ 5/5 182.46 0.05 n/a
mem_copy_4_cores_1_chans_2048_tile_512_False ✅ 5/5 190.84 0.04 n/a
mem_copy_4_cores_2_chans_2048_tile_512_False ✅ 5/5 192.50 0.05 n/a
mem_copy_8_cores_1_chans_2048_tile_256_False ✅ 5/5 189.56 0.04 n/a
mem_copy_8_cores_2_chans_2048_tile_256_False ✅ 5/5 221.36 0.04 n/a
mha ✅ 5/5 47476.36 0.18 n/a
relu_1_cols_1_channels_2048_tile_2048 ✅ 5/5 191.58 0.04 n/a
relu_2_cols_1_channels_2048_tile_1024 ✅ 5/5 197.48 0.04 n/a
relu_4_cols_1_channels_2048_tile_512 ✅ 5/5 195.32 0.04 n/a
relu_8_cols_1_channels_2048_tile_256 ✅ 5/5 197.56 0.04 n/a
rms_norm_1_cols_1_channels_2048_tile_2048 ✅ 5/5 201.72 0.05 n/a
rms_norm_1_cols_2_channels_2048_tile_1024 ✅ 5/5 133.46 0.07 n/a
rms_norm_2_cols_1_channels_2048_tile_1024 ✅ 5/5 189.92 0.04 n/a
rms_norm_2_cols_2_channels_2048_tile_512 ✅ 5/5 166.40 0.05 n/a
rms_norm_4_cols_1_channels_2048_tile_512 ✅ 5/5 181.98 0.05 n/a
rms_norm_4_cols_2_channels_2048_tile_256 ✅ 5/5 226.64 0.04 n/a
rms_norm_8_cols_1_channels_2048_tile_256 ✅ 5/5 256.68 0.03 n/a
rms_norm_8_cols_2_channels_2048_tile_128 ✅ 5/5 262.08 0.03 n/a
rope_1c_32rows_512cols_32arows_0m ✅ 5/5 159.58 0.67 n/a
rope_1c_32rows_512cols_8arows_0m ✅ 5/5 190.32 0.42 n/a
rope_2c_32rows_512cols_32arows_0m ✅ 5/5 168.20 0.61 n/a
rope_2c_32rows_512cols_8arows_0m ✅ 5/5 169.24 0.45 n/a
rope_8c_32rows_512cols_32arows_0m ✅ 5/5 223.52 0.45 n/a
rope_8c_32rows_512cols_8arows_0m ✅ 5/5 176.94 0.43 n/a
sigmoid_1_cols_1_channels_2048_tile_2048 ✅ 5/5 162.56 0.05 n/a
sigmoid_2_cols_1_channels_2048_tile_1024 ✅ 5/5 167.02 0.05 n/a
sigmoid_4_cols_1_channels_2048_tile_512 ✅ 5/5 203.60 0.04 n/a
sigmoid_8_cols_1_channels_2048_tile_256 ✅ 5/5 172.28 0.05 n/a
silu_1_cols_1_channels_2048_tile_2048 ✅ 5/5 190.48 0.04 n/a
silu_2_cols_1_channels_2048_tile_1024 ✅ 5/5 201.48 0.04 n/a
silu_4_cols_1_channels_2048_tile_512 ✅ 5/5 188.20 0.05 n/a
silu_8_cols_1_channels_2048_tile_256 ✅ 5/5 222.76 0.04 n/a
softmax_2_cols_2_channels_32768_tile_1024 ✅ 5/5 164.16 0.83 n/a
softmax_2_cols_2_channels_32768_tile_512 ✅ 5/5 155.98 0.91 n/a
softmax_4_cols_4_channels_32768_tile_2048 ✅ 5/5 262.78 0.51 n/a
swiglu_decode_1x2048x2048 ✅ 5/5 n/a n/a n/a
swiglu_prefill_256x2048x2048 ✅ 5/5 n/a n/a n/a
tanh_1_cols_1_channels_2048_tile_2048 ✅ 5/5 175.70 0.05 n/a
tanh_2_cols_1_channels_2048_tile_1024 ✅ 5/5 171.10 0.05 n/a
tanh_4_cols_1_channels_2048_tile_512 ✅ 5/5 173.48 0.05 n/a
tanh_8_cols_1_channels_2048_tile_256 ✅ 5/5 206.98 0.04 n/a
transpose_2048_M_64_N_1_cols_1_channels_64_m_64_n_8_s ✅ 5/5 216.18 2.52 n/a
transpose_2048_M_64_N_1_cols_2_channels_64_m_64_n_8_s ✅ 5/5 162.66 3.35 n/a
weighted_rms_norm_1_cols_2_channels_2048_weights_2048 ✅ 5/5 186.50 0.07 n/a
weighted_rms_norm_2_cols_2_channels_2048_weights_1024 ✅ 5/5 207.74 0.05 n/a
weighted_rms_norm_4_cols_2_channels_2048_weights_512 ✅ 5/5 217.16 0.04 n/a
weighted_rms_norm_8_cols_2_channels_2048_weights_256 ✅ 5/5 230.78 0.04 n/a
📈 Trends (vs main branch) for Small Benchmark/Test Suite

e73d4ff (2026_03_20_22_10_55)

IRONCLAD Trends

axpy_1_cols_2_channels_2048_tile_2048_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.09 (-30.93%)0.07 (-31.28%)0.07 (-34.11%)0.06 (-26.59%)0.01 (-24.87%)221.30 (+36.27%)175.30 (+44.92%)177.60 (+51.79%)136.60 (+44.70%)33.35 (+44.30%)
130b6ea — 2025-12-05 21:15:490.13 (n/a)0.11 (n/a)0.10 (n/a)0.08 (n/a)0.02 (n/a)162.40 (n/a)120.96 (n/a)117.00 (n/a)94.40 (n/a)23.11 (n/a)

axpy_1_cols_2_channels_2048_tile_2048_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.16 (-3.31%)0.14 (-1.29%)0.15 (+16.18%)0.10 (-6.80%)0.03 (+3.89%)125.70 (+7.25%)93.56 (+2.12%)83.50 (-13.92%)75.00 (+3.45%)22.38 (+18.09%)
84d3478 — 2026-02-17 23:25:130.17 (n/a)0.14 (n/a)0.13 (n/a)0.10 (n/a)0.03 (n/a)117.20 (n/a)91.62 (n/a)97.00 (n/a)72.50 (n/a)18.95 (n/a)

axpy_2_cols_2_channels_2048_tile_1024_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.07 (-35.86%)0.06 (-35.01%)0.06 (-29.21%)0.05 (-38.04%)0.01 (-25.02%)245.30 (+61.38%)208.80 (+54.14%)201.70 (+41.25%)179.30 (+55.91%)28.55 (+89.92%)
130b6ea — 2025-12-05 21:15:490.11 (n/a)0.09 (n/a)0.09 (n/a)0.08 (n/a)0.01 (n/a)152.00 (n/a)135.46 (n/a)142.80 (n/a)115.00 (n/a)15.03 (n/a)

axpy_2_cols_2_channels_2048_tile_1024_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (-0.73%)0.11 (-7.43%)0.11 (-8.63%)0.07 (-27.71%)0.02 (+50.39%)182.70 (+38.30%)120.86 (+11.97%)110.80 (+9.38%)96.10 (+0.73%)35.16 (+122.88%)
84d3478 — 2026-02-17 23:25:130.13 (n/a)0.12 (n/a)0.12 (n/a)0.09 (n/a)0.02 (n/a)132.10 (n/a)107.94 (n/a)101.30 (n/a)95.40 (n/a)15.78 (n/a)

axpy_4_cols_2_channels_2048_tile_512_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.09 (-20.74%)0.07 (-27.00%)0.07 (-33.84%)0.05 (-27.26%)0.02 (+6.40%)229.10 (+37.43%)179.26 (+39.20%)184.10 (+51.15%)131.10 (+26.18%)41.87 (+82.47%)
130b6ea — 2025-12-05 21:15:490.12 (n/a)0.10 (n/a)0.10 (n/a)0.07 (n/a)0.02 (n/a)166.70 (n/a)128.78 (n/a)121.80 (n/a)103.90 (n/a)22.95 (n/a)

axpy_4_cols_2_channels_2048_tile_512_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (-10.91%)0.11 (-7.68%)0.12 (-0.73%)0.06 (-17.94%)0.03 (+0.99%)194.10 (+21.85%)124.52 (+10.29%)103.30 (+0.68%)97.20 (+12.24%)40.32 (+39.15%)
84d3478 — 2026-02-17 23:25:130.14 (n/a)0.11 (n/a)0.12 (n/a)0.08 (n/a)0.03 (n/a)159.30 (n/a)112.90 (n/a)102.60 (n/a)86.60 (n/a)28.97 (n/a)

axpy_8_cols_2_channels_2048_tile_256_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.07 (-23.96%)0.06 (-23.57%)0.06 (-28.24%)0.06 (-17.46%)0.01 (-37.40%)223.40 (+21.15%)198.44 (+29.55%)200.70 (+39.28%)167.10 (+31.47%)20.67 (-2.79%)
130b6ea — 2025-12-05 21:15:490.10 (n/a)0.08 (n/a)0.09 (n/a)0.07 (n/a)0.01 (n/a)184.40 (n/a)153.18 (n/a)144.10 (n/a)127.10 (n/a)21.26 (n/a)

axpy_8_cols_2_channels_2048_tile_256_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (+0.01%)0.11 (+6.10%)0.11 (+0.22%)0.08 (-0.33%)0.02 (-9.83%)156.30 (+0.32%)115.26 (-6.40%)112.70 (-0.18%)95.50 (+0.00%)24.34 (-10.22%)
84d3478 — 2026-02-17 23:25:130.13 (n/a)0.10 (n/a)0.11 (n/a)0.08 (n/a)0.02 (n/a)155.80 (n/a)123.14 (n/a)112.90 (n/a)95.50 (n/a)27.11 (n/a)

dequant_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.04 (-4.71%)0.03 (-18.79%)0.03 (-24.32%)0.02 (-32.71%)0.01 (+75.48%)238.40 (+48.63%)181.14 (+28.36%)180.20 (+32.11%)119.90 (+4.90%)46.36 (+166.93%)
130b6ea — 2025-12-05 21:15:490.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.00 (n/a)160.40 (n/a)141.12 (n/a)136.40 (n/a)114.30 (n/a)17.37 (n/a)

dequant_1_cols_1_channels_2048_tile_2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-13.00%)0.04 (-0.87%)0.04 (+1.16%)0.03 (-3.17%)0.01 (-31.33%)158.40 (+3.26%)128.12 (-0.19%)123.80 (-1.12%)110.40 (+14.88%)18.36 (-15.65%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)153.40 (n/a)128.36 (n/a)125.20 (n/a)96.10 (n/a)21.77 (n/a)

dequant_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.05 (-22.06%)0.03 (-31.35%)0.04 (-34.04%)0.02 (-44.85%)0.01 (+17.81%)242.20 (+81.29%)160.98 (+51.75%)144.20 (+51.63%)112.80 (+28.33%)50.24 (+180.69%)
130b6ea — 2025-12-05 21:15:490.06 (n/a)0.05 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)133.60 (n/a)106.08 (n/a)95.10 (n/a)87.90 (n/a)17.90 (n/a)

dequant_1_cols_2_channels_2048_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (-3.74%)0.05 (-12.63%)0.05 (-5.21%)0.03 (-33.47%)0.01 (+52.60%)162.70 (+50.37%)110.84 (+18.98%)98.80 (+5.44%)81.50 (+3.95%)31.02 (+149.23%)
84d3478 — 2026-02-17 23:25:130.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)108.20 (n/a)93.16 (n/a)93.70 (n/a)78.40 (n/a)12.45 (n/a)

dequant_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.05 (-31.90%)0.04 (-32.54%)0.04 (-27.67%)0.03 (-29.25%)0.01 (-13.77%)202.70 (+41.35%)151.02 (+49.50%)134.90 (+38.22%)115.90 (+46.89%)39.05 (+73.06%)
130b6ea — 2025-12-05 21:15:490.07 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)143.40 (n/a)101.02 (n/a)97.60 (n/a)78.90 (n/a)22.56 (n/a)

dequant_2_cols_1_channels_2048_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (+8.90%)0.05 (+1.31%)0.05 (-1.85%)0.03 (+6.35%)0.01 (+7.11%)152.30 (-5.93%)115.08 (-1.20%)100.80 (+1.92%)88.10 (-8.23%)27.20 (-4.42%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)161.90 (n/a)116.48 (n/a)98.90 (n/a)96.00 (n/a)28.46 (n/a)

dequant_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.04 (-24.59%)0.03 (-29.63%)0.03 (-36.65%)0.02 (-28.07%)0.01 (-25.93%)222.00 (+39.01%)178.58 (+41.37%)172.10 (+57.89%)141.60 (+32.58%)31.37 (+39.53%)
130b6ea — 2025-12-05 21:15:490.05 (n/a)0.04 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)159.70 (n/a)126.32 (n/a)109.00 (n/a)106.80 (n/a)22.48 (n/a)

dequant_2_cols_2_channels_2048_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.07 (+27.96%)0.05 (+1.57%)0.04 (-18.06%)0.03 (+10.23%)0.01 (+22.79%)164.70 (-9.31%)119.12 (-1.39%)120.20 (+22.03%)75.60 (-21.82%)31.61 (-13.80%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)181.60 (n/a)120.80 (n/a)98.50 (n/a)96.70 (n/a)36.67 (n/a)

dequant_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.04 (-35.48%)0.03 (-38.15%)0.03 (-41.22%)0.02 (-31.45%)0.01 (-33.66%)227.70 (+45.87%)193.62 (+60.20%)195.50 (+70.15%)136.90 (+55.04%)35.36 (+44.92%)
130b6ea — 2025-12-05 21:15:490.06 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)156.10 (n/a)120.86 (n/a)114.90 (n/a)88.30 (n/a)24.40 (n/a)

dequant_4_cols_1_channels_2048_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-2.55%)0.04 (-12.79%)0.05 (-16.25%)0.03 (-12.85%)0.01 (-1.20%)164.60 (+14.78%)122.14 (+15.14%)115.90 (+19.48%)97.50 (+2.63%)25.08 (+19.84%)
84d3478 — 2026-02-17 23:25:130.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)143.40 (n/a)106.08 (n/a)97.00 (n/a)95.00 (n/a)20.93 (n/a)

dequant_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.04 (-16.56%)0.03 (-22.58%)0.03 (-25.28%)0.02 (-22.76%)0.01 (-6.98%)217.70 (+29.51%)175.50 (+29.29%)176.90 (+33.81%)128.70 (+19.83%)32.21 (+40.64%)
130b6ea — 2025-12-05 21:15:490.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)168.10 (n/a)135.74 (n/a)132.20 (n/a)107.40 (n/a)22.91 (n/a)

dequant_4_cols_2_channels_2048_tile_256_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (+20.83%)0.05 (+8.21%)0.04 (+1.42%)0.03 (+17.39%)0.01 (+14.84%)166.40 (-14.80%)122.34 (-7.97%)127.70 (-1.39%)84.60 (-17.22%)30.47 (-19.21%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)195.30 (n/a)132.94 (n/a)129.50 (n/a)102.20 (n/a)37.71 (n/a)

dequant_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.04 (-12.27%)0.03 (-28.06%)0.02 (-39.10%)0.02 (-30.95%)0.01 (+58.40%)246.80 (+44.84%)201.66 (+43.92%)222.90 (+64.14%)136.40 (+14.05%)46.83 (+161.03%)
130b6ea — 2025-12-05 21:15:490.04 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.00 (n/a)170.40 (n/a)140.12 (n/a)135.80 (n/a)119.60 (n/a)17.94 (n/a)

dequant_8_cols_1_channels_2048_tile_256_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-25.19%)0.04 (-21.02%)0.04 (-19.70%)0.04 (-19.76%)0.00 (-37.92%)142.70 (+24.63%)126.44 (+26.24%)125.70 (+24.46%)115.50 (+33.68%)10.61 (+3.97%)
84d3478 — 2026-02-17 23:25:130.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)114.50 (n/a)100.16 (n/a)101.00 (n/a)86.40 (n/a)10.20 (n/a)

dequant_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.03 (-43.25%)0.02 (-28.89%)0.02 (-26.31%)0.02 (-18.83%)0.00 (-60.87%)244.10 (+23.16%)218.22 (+36.39%)225.50 (+35.68%)187.60 (+76.15%)26.69 (-10.62%)
130b6ea — 2025-12-05 21:15:490.05 (n/a)0.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)198.20 (n/a)160.00 (n/a)166.20 (n/a)106.50 (n/a)29.86 (n/a)

dequant_8_cols_2_channels_2048_tile_128_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.04 (-26.69%)0.04 (-11.93%)0.04 (-16.99%)0.03 (+27.28%)0.01 (-62.60%)193.10 (-21.41%)147.68 (+2.96%)136.10 (+20.44%)135.40 (+36.49%)25.41 (-58.96%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.05 (n/a)0.02 (n/a)0.01 (n/a)245.70 (n/a)143.44 (n/a)113.00 (n/a)99.20 (n/a)61.90 (n/a)

eltwise_add_1_cols_2_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.08 (-50.48%)0.07 (-48.43%)0.07 (-53.64%)0.05 (-37.30%)0.01 (-69.10%)232.50 (+59.57%)185.30 (+86.64%)180.00 (+115.83%)156.60 (+101.80%)29.21 (+1.79%)
897d04e — 2026-03-06 22:12:250.16 (n/a)0.13 (n/a)0.15 (n/a)0.08 (n/a)0.03 (n/a)145.70 (n/a)99.28 (n/a)83.40 (n/a)77.60 (n/a)28.70 (n/a)

eltwise_add_2_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.07 (-42.69%)0.07 (-36.55%)0.07 (-43.55%)0.06 (-19.71%)0.00 (-85.65%)196.90 (+24.54%)186.26 (+52.02%)186.00 (+77.14%)176.90 (+74.46%)8.71 (-68.35%)
897d04e — 2026-03-06 22:12:250.12 (n/a)0.10 (n/a)0.12 (n/a)0.08 (n/a)0.02 (n/a)158.10 (n/a)122.52 (n/a)105.00 (n/a)101.40 (n/a)27.53 (n/a)

eltwise_add_4_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.10 (-38.50%)0.07 (-39.98%)0.07 (-49.42%)0.06 (-10.79%)0.02 (-62.15%)222.70 (+12.08%)176.58 (+52.57%)170.90 (+97.57%)129.00 (+62.67%)35.30 (-30.38%)
897d04e — 2026-03-06 22:12:250.15 (n/a)0.12 (n/a)0.14 (n/a)0.06 (n/a)0.04 (n/a)198.70 (n/a)115.74 (n/a)86.50 (n/a)79.30 (n/a)50.70 (n/a)

eltwise_add_8_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.09 (-22.88%)0.07 (-30.63%)0.07 (-34.41%)0.06 (-17.40%)0.01 (-34.91%)213.10 (+21.01%)178.48 (+42.62%)171.90 (+52.53%)144.50 (+29.71%)29.33 (+2.89%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.10 (n/a)0.11 (n/a)0.07 (n/a)0.02 (n/a)176.10 (n/a)125.14 (n/a)112.70 (n/a)111.40 (n/a)28.50 (n/a)

eltwise_mul_1_cols_2_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.09 (-39.98%)0.08 (-40.94%)0.08 (-44.81%)0.06 (-16.48%)0.01 (-62.42%)198.80 (+19.69%)165.50 (+61.18%)161.60 (+81.17%)139.50 (+66.67%)25.72 (-27.69%)
897d04e — 2026-03-06 22:12:250.15 (n/a)0.13 (n/a)0.14 (n/a)0.07 (n/a)0.03 (n/a)166.10 (n/a)102.68 (n/a)89.20 (n/a)83.70 (n/a)35.57 (n/a)

eltwise_mul_2_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.10 (-31.40%)0.08 (-40.48%)0.09 (-42.42%)0.05 (-42.87%)0.02 (-23.53%)235.30 (+75.07%)164.54 (+71.25%)143.10 (+73.67%)118.70 (+45.64%)46.14 (+102.04%)
897d04e — 2026-03-06 22:12:250.15 (n/a)0.13 (n/a)0.15 (n/a)0.09 (n/a)0.03 (n/a)134.40 (n/a)96.08 (n/a)82.40 (n/a)81.50 (n/a)22.84 (n/a)

eltwise_mul_4_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.06 (-44.25%)0.05 (-49.38%)0.06 (-45.96%)0.03 (-51.77%)0.02 (-29.51%)407.20 (+107.33%)263.78 (+105.12%)213.60 (+85.10%)190.00 (+79.41%)95.13 (+148.21%)
897d04e — 2026-03-06 22:12:250.12 (n/a)0.10 (n/a)0.11 (n/a)0.06 (n/a)0.02 (n/a)196.40 (n/a)128.60 (n/a)115.40 (n/a)105.90 (n/a)38.33 (n/a)

eltwise_mul_8_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.06 (-43.49%)0.05 (-48.07%)0.05 (-51.08%)0.04 (-54.59%)0.01 (-21.59%)341.90 (+120.30%)242.72 (+96.79%)232.50 (+104.49%)198.60 (+77.01%)57.56 (+213.60%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.10 (n/a)0.11 (n/a)0.08 (n/a)0.01 (n/a)155.20 (n/a)123.34 (n/a)113.70 (n/a)112.20 (n/a)18.35 (n/a)

gelu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.06 (-47.16%)0.04 (-51.97%)0.04 (-58.53%)0.03 (-45.48%)0.01 (-39.85%)247.50 (+83.47%)199.20 (+109.95%)212.00 (+141.18%)146.90 (+89.30%)46.77 (+100.74%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.09 (n/a)0.06 (n/a)0.02 (n/a)134.90 (n/a)94.88 (n/a)87.90 (n/a)77.60 (n/a)23.30 (n/a)

gelu_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.05 (-52.42%)0.04 (-56.30%)0.04 (-62.47%)0.03 (-55.37%)0.01 (-39.01%)249.60 (+124.06%)196.44 (+132.20%)205.80 (+166.58%)152.80 (+110.18%)42.08 (+168.76%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.10 (n/a)0.11 (n/a)0.07 (n/a)0.02 (n/a)111.40 (n/a)84.60 (n/a)77.20 (n/a)72.70 (n/a)15.66 (n/a)

gelu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.06 (-43.78%)0.05 (-48.70%)0.05 (-54.81%)0.03 (-34.81%)0.01 (-58.48%)239.10 (+53.37%)178.28 (+87.47%)175.70 (+121.28%)136.40 (+77.84%)38.43 (+12.43%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.10 (n/a)0.05 (n/a)0.02 (n/a)155.90 (n/a)95.10 (n/a)79.40 (n/a)76.70 (n/a)34.18 (n/a)

gelu_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.06 (-43.90%)0.05 (-49.49%)0.05 (-55.95%)0.03 (-37.24%)0.01 (-58.12%)241.80 (+59.39%)176.86 (+90.91%)173.80 (+126.89%)133.50 (+78.24%)39.92 (+20.10%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.10 (n/a)0.11 (n/a)0.05 (n/a)0.02 (n/a)151.70 (n/a)92.64 (n/a)76.60 (n/a)74.90 (n/a)33.24 (n/a)

gelu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.07 (-38.45%)0.05 (-47.93%)0.05 (-52.12%)0.03 (-44.34%)0.01 (-34.95%)271.80 (+79.64%)184.88 (+94.28%)175.00 (+108.83%)121.40 (+62.52%)58.98 (+84.49%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.10 (n/a)0.05 (n/a)0.02 (n/a)151.30 (n/a)95.16 (n/a)83.80 (n/a)74.70 (n/a)31.97 (n/a)

gelu_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.07 (-33.61%)0.05 (-45.81%)0.04 (-53.83%)0.04 (-15.88%)0.01 (-51.52%)215.80 (+18.90%)182.48 (+74.79%)183.80 (+116.49%)124.90 (+50.66%)36.26 (-16.00%)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.09 (n/a)0.10 (n/a)0.05 (n/a)0.02 (n/a)181.50 (n/a)104.40 (n/a)84.90 (n/a)82.90 (n/a)43.16 (n/a)

gelu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.06 (-40.71%)0.04 (-49.04%)0.04 (-57.05%)0.03 (-41.88%)0.01 (-30.64%)246.50 (+72.02%)198.58 (+99.02%)212.70 (+132.97%)137.60 (+68.63%)48.78 (+95.47%)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.09 (n/a)0.09 (n/a)0.06 (n/a)0.02 (n/a)143.30 (n/a)99.78 (n/a)91.30 (n/a)81.60 (n/a)24.96 (n/a)

gelu_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.04 (-47.80%)0.03 (-54.01%)0.03 (-56.16%)0.03 (-45.89%)0.01 (-51.22%)320.90 (+84.74%)252.14 (+115.73%)242.20 (+128.06%)188.10 (+91.55%)54.40 (+69.89%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.07 (n/a)0.08 (n/a)0.05 (n/a)0.01 (n/a)173.70 (n/a)116.88 (n/a)106.20 (n/a)98.20 (n/a)32.02 (n/a)

gemm_1792x896x1152_64x32x48_8cols_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
e73d4ff — 2026-03-20 22:04:504.89 (-2.13%)4.38 (-10.35%)4.17 (-16.00%)4.04 (-10.47%)0.40 (+93.43%)2327.20 (+11.69%)2162.06 (+12.11%)2254.20 (+19.04%)1923.30 (+2.18%)191.70 (+120.92%)1923.43 (-2.13%)1722.22 (-10.35%)1641.07 (-16.00%)1589.62 (-10.47%)157.78 (+93.43%)
897d04e — 2026-03-06 22:12:255.00 (n/a)4.88 (n/a)4.97 (n/a)4.51 (n/a)0.21 (n/a)2083.60 (n/a)1928.60 (n/a)1893.70 (n/a)1882.30 (n/a)86.77 (n/a)1965.38 (n/a)1921.12 (n/a)1953.55 (n/a)1775.47 (n/a)81.57 (n/a)

gemm_192x384x64_48x96x16_4cols

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
e73d4ff — 2026-03-20 22:04:501.06 (-48.37%)0.88 (-47.11%)0.95 (-48.97%)0.47 (-40.42%)0.24 (-54.44%)472.50 (+67.79%)276.14 (+82.01%)232.50 (+96.04%)209.10 (+93.79%)111.13 (+50.37%)45.13 (-48.37%)37.44 (-47.11%)40.60 (-48.97%)19.97 (-40.42%)10.24 (-54.44%)
897d04e — 2026-03-06 22:12:252.05 (n/a)1.66 (n/a)1.86 (n/a)0.79 (n/a)0.53 (n/a)281.60 (n/a)151.72 (n/a)118.60 (n/a)107.90 (n/a)73.91 (n/a)87.43 (n/a)70.78 (n/a)79.55 (n/a)33.52 (n/a)22.49 (n/a)

gemm_192x384x64_48x96x16_4cols_bcolmaj_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
e73d4ff — 2026-03-20 22:04:501.11 (-50.35%)0.84 (-53.93%)0.87 (-53.63%)0.49 (-63.16%)0.24 (-27.03%)455.60 (+171.51%)285.48 (+128.71%)255.10 (+115.64%)199.60 (+101.41%)101.41 (+296.48%)47.29 (-50.35%)35.87 (-53.93%)36.99 (-53.63%)20.71 (-63.16%)10.19 (-27.03%)
897d04e — 2026-03-06 22:12:252.23 (n/a)1.82 (n/a)1.87 (n/a)1.32 (n/a)0.33 (n/a)167.80 (n/a)124.82 (n/a)118.30 (n/a)99.10 (n/a)25.58 (n/a)95.25 (n/a)77.86 (n/a)79.77 (n/a)56.23 (n/a)13.96 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.72 (-0.11%)3.55 (-2.75%)3.56 (-2.75%)3.36 (-5.33%)0.12 (+76.10%)2725.23 (-0.11%)2602.52 (-2.75%)2612.11 (-2.75%)2459.89 (-5.33%)86.27 (+76.10%)6984.00 (+5.63%)6608.60 (+2.91%)6577.00 (+2.83%)6304.00 (+0.11%)221.87 (+87.15%)
9e98049 — 2025-11-25 20:50:553.72 (n/a)3.65 (n/a)3.66 (n/a)3.54 (n/a)0.07 (n/a)2728.26 (n/a)2676.06 (n/a)2686.03 (n/a)2598.29 (n/a)48.99 (n/a)6612.00 (n/a)6422.00 (n/a)6396.00 (n/a)6297.00 (n/a)118.55 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_0_bcolmaj_1_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.40 (-3.14%)3.38 (-0.82%)3.38 (-0.82%)3.35 (+4.93%)0.02 (-84.51%)2493.81 (-3.14%)2475.27 (-0.82%)2479.06 (-0.82%)2453.92 (+4.93%)13.33 (-84.51%)7001.00 (-4.70%)6940.80 (+0.71%)6930.00 (+0.83%)6889.00 (+3.24%)37.43 (-84.76%)
9e98049 — 2025-11-25 20:50:553.51 (n/a)3.40 (n/a)3.41 (n/a)3.19 (n/a)0.12 (n/a)2574.53 (n/a)2495.72 (n/a)2499.62 (n/a)2338.67 (n/a)86.02 (n/a)7346.00 (n/a)6892.20 (n/a)6873.00 (n/a)6673.00 (n/a)245.57 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_1_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.02 (+0.52%)2.87 (-2.05%)2.86 (-2.24%)2.70 (-6.16%)0.11 (+131.66%)2216.19 (+0.52%)2104.44 (-2.05%)2094.08 (-2.24%)1980.84 (-6.16%)80.25 (+131.67%)8673.00 (+6.56%)8175.60 (+2.21%)8204.00 (+2.29%)7752.00 (-0.51%)314.03 (+145.36%)
9e98049 — 2025-11-25 20:50:553.01 (n/a)2.93 (n/a)2.92 (n/a)2.88 (n/a)0.05 (n/a)2204.81 (n/a)2148.41 (n/a)2142.13 (n/a)2110.81 (n/a)34.64 (n/a)8139.00 (n/a)7998.60 (n/a)8020.00 (n/a)7792.00 (n/a)127.99 (n/a)

gemm_2048x2048x2048_64x64x64_1cols

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
e73d4ff — 2026-03-20 22:04:500.51 (-2.68%)0.51 (-2.19%)0.51 (-1.94%)0.51 (-2.30%)0.00 (-51.29%)49539.30 (+2.35%)49398.32 (+2.23%)49346.90 (+1.98%)49335.50 (+2.75%)87.49 (-48.78%)348.23 (-2.68%)347.78 (-2.19%)348.14 (-1.94%)346.79 (-2.30%)0.62 (-51.29%)
897d04e — 2026-03-06 22:12:250.52 (n/a)0.52 (n/a)0.52 (n/a)0.52 (n/a)0.00 (n/a)48401.80 (n/a)48318.62 (n/a)48391.10 (n/a)48013.20 (n/a)170.80 (n/a)357.82 (n/a)355.56 (n/a)355.02 (n/a)354.94 (n/a)1.26 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:491.04 (+0.94%)1.03 (+0.23%)1.03 (+0.09%)1.02 (+0.07%)0.00 (+197.51%)24576.00 (-0.07%)24485.82 (-0.23%)24530.40 (-0.09%)24266.40 (-0.94%)111.26 (+194.35%)
0a6c11c — 2025-12-04 00:39:101.03 (n/a)1.03 (n/a)1.03 (n/a)1.02 (n/a)0.00 (n/a)24594.30 (n/a)24542.14 (n/a)24551.90 (n/a)24495.50 (n/a)37.80 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:051.04 (+0.28%)1.03 (-0.15%)1.03 (+0.05%)1.02 (-0.52%)0.01 (+42.72%)24719.80 (+0.52%)24512.42 (+0.16%)24529.20 (-0.05%)24092.60 (-0.28%)252.92 (+43.15%)713.08 (+0.28%)700.92 (-0.15%)700.38 (+0.05%)694.98 (-0.52%)7.30 (+42.71%)
50ad37c — 2025-12-29 20:35:531.04 (n/a)1.03 (n/a)1.03 (n/a)1.02 (n/a)0.01 (n/a)24592.30 (n/a)24473.70 (n/a)24540.90 (n/a)24160.80 (n/a)176.68 (n/a)711.06 (n/a)702.00 (n/a)700.05 (n/a)698.59 (n/a)5.12 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_1_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:490.99 (+1.69%)0.97 (+0.29%)0.97 (-0.04%)0.97 (-0.15%)0.01 (+434.92%)26025.30 (+0.15%)25839.34 (-0.29%)25916.70 (+0.04%)25446.00 (-1.67%)202.05 (+425.44%)
0a6c11c — 2025-12-04 00:39:100.97 (n/a)0.97 (n/a)0.97 (n/a)0.97 (n/a)0.00 (n/a)25986.90 (n/a)25913.52 (n/a)25905.70 (n/a)25877.00 (n/a)38.45 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_1_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:050.97 (-0.01%)0.97 (+0.03%)0.97 (+0.04%)0.97 (+0.00%)0.00 (-6.43%)25927.90 (-0.00%)25903.20 (-0.03%)25898.60 (-0.04%)25892.80 (+0.01%)14.05 (-6.56%)663.50 (-0.01%)663.23 (+0.03%)663.35 (+0.04%)662.60 (+0.00%)0.36 (-6.43%)
50ad37c — 2025-12-29 20:35:530.97 (n/a)0.97 (n/a)0.97 (n/a)0.97 (n/a)0.00 (n/a)25928.10 (n/a)25910.14 (n/a)25908.10 (n/a)25891.20 (n/a)15.04 (n/a)663.54 (n/a)663.06 (n/a)663.11 (n/a)662.60 (n/a)0.38 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_1_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:490.92 (-1.02%)0.92 (-0.66%)0.92 (-0.10%)0.91 (-0.78%)0.01 (-4.54%)27751.40 (+0.79%)27480.74 (+0.66%)27354.80 (+0.10%)27320.50 (+1.03%)182.69 (-2.84%)
0a6c11c — 2025-12-04 00:39:100.93 (n/a)0.92 (n/a)0.92 (n/a)0.91 (n/a)0.01 (n/a)27534.80 (n/a)27299.32 (n/a)27326.70 (n/a)27042.30 (n/a)188.04 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_1_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:050.92 (+0.11%)0.91 (-0.12%)0.91 (-0.21%)0.91 (-0.24%)0.00 (+30.34%)27723.50 (+0.24%)27548.94 (+0.12%)27522.90 (+0.21%)27403.60 (-0.11%)124.26 (+30.58%)626.92 (+0.11%)623.62 (-0.12%)624.20 (-0.21%)619.69 (-0.24%)2.81 (+30.34%)
50ad37c — 2025-12-29 20:35:530.92 (n/a)0.91 (n/a)0.92 (n/a)0.91 (n/a)0.00 (n/a)27657.10 (n/a)27515.58 (n/a)27464.70 (n/a)27434.60 (n/a)95.16 (n/a)626.21 (n/a)624.37 (n/a)625.52 (n/a)621.17 (n/a)2.16 (n/a)

gemm_2048x2048x2048_64x64x64_2cols_bcolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
e73d4ff — 2026-03-20 22:04:500.88 (-3.70%)0.88 (-2.85%)0.88 (-3.43%)0.87 (-1.33%)0.00 (-67.61%)28830.10 (+1.35%)28712.80 (+2.92%)28760.90 (+3.55%)28577.80 (+3.85%)122.41 (-65.96%)601.16 (-3.70%)598.34 (-2.85%)597.33 (-3.43%)595.90 (-1.33%)2.55 (-67.61%)
897d04e — 2026-03-06 22:12:250.91 (n/a)0.90 (n/a)0.91 (n/a)0.88 (n/a)0.01 (n/a)28447.10 (n/a)27899.18 (n/a)27774.40 (n/a)27519.50 (n/a)359.64 (n/a)624.28 (n/a)615.87 (n/a)618.55 (n/a)603.92 (n/a)7.88 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
31ab24e — 2025-11-22 02:01:563.65 (-1.06%)3.59 (+0.83%)3.63 (-0.26%)3.45 (+2.66%)0.07 (-44.71%)2678.08 (-1.06%)2632.13 (+0.83%)2658.60 (-0.26%)2528.31 (+2.66%)53.86 (-44.71%)6795.00 (-2.59%)6529.80 (-0.92%)6462.00 (+0.26%)6415.00 (+1.07%)137.14 (-45.27%)
d9a64c5 — 2025-11-21 20:49:353.69 (n/a)3.56 (n/a)3.64 (n/a)3.36 (n/a)0.13 (n/a)2706.77 (n/a)2610.50 (n/a)2665.61 (n/a)2462.71 (n/a)97.41 (n/a)6976.00 (n/a)6590.40 (n/a)6445.00 (n/a)6347.00 (n/a)250.58 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.98 (+0.20%)3.84 (-2.95%)3.76 (-5.27%)3.72 (-4.97%)0.13 (+473.97%)6760.00 (+5.23%)6559.50 (+3.13%)6697.10 (+5.56%)6321.00 (-0.20%)218.68 (+499.40%)2717.91 (+0.20%)2621.44 (-2.95%)2565.27 (-5.27%)2541.40 (-4.97%)88.37 (+473.97%)
331dcca — 2026-01-08 18:34:103.97 (n/a)3.96 (n/a)3.97 (n/a)3.92 (n/a)0.02 (n/a)6423.80 (n/a)6360.32 (n/a)6344.30 (n/a)6333.90 (n/a)36.48 (n/a)2712.39 (n/a)2701.17 (n/a)2707.91 (n/a)2674.40 (n/a)15.40 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_1_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.75 (+0.28%)3.70 (+1.25%)3.71 (-0.21%)3.65 (+4.77%)0.04 (-66.90%)6901.80 (-4.56%)6795.46 (-1.30%)6790.20 (+0.21%)6715.90 (-0.28%)67.05 (-68.40%)2558.08 (+0.28%)2528.33 (+1.25%)2530.10 (-0.21%)2489.18 (+4.77%)24.81 (-66.90%)
331dcca — 2026-01-08 18:34:103.74 (n/a)3.66 (n/a)3.71 (n/a)3.48 (n/a)0.11 (n/a)7231.20 (n/a)6885.18 (n/a)6775.70 (n/a)6734.90 (n/a)212.21 (n/a)2550.87 (n/a)2497.04 (n/a)2535.50 (n/a)2375.78 (n/a)74.96 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_1_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.62 (-0.77%)3.54 (+0.14%)3.60 (-0.04%)3.44 (+2.17%)0.09 (-25.77%)7307.40 (-2.13%)7103.46 (-0.18%)6982.50 (+0.04%)6948.20 (+0.77%)182.95 (-26.55%)2472.55 (-0.77%)2419.80 (+0.14%)2460.41 (-0.04%)2351.04 (+2.17%)61.76 (-25.77%)
331dcca — 2026-01-08 18:34:103.65 (n/a)3.54 (n/a)3.61 (n/a)3.37 (n/a)0.12 (n/a)7466.20 (n/a)7116.40 (n/a)6979.60 (n/a)6895.10 (n/a)249.09 (n/a)2491.61 (n/a)2416.45 (n/a)2461.45 (n/a)2301.00 (n/a)83.20 (n/a)

gemm_2048x2048x2048_64x64x64_8cols_bcolmaj_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
e73d4ff — 2026-03-20 22:04:503.45 (-1.55%)3.31 (-5.04%)3.31 (-4.91%)3.12 (-10.27%)0.13 (+942.73%)8063.60 (+11.45%)7607.48 (+5.44%)7600.10 (+5.16%)7284.60 (+1.57%)301.54 (+1081.86%)2358.37 (-1.55%)2261.08 (-5.04%)2260.49 (-4.91%)2130.56 (-10.27%)88.08 (+942.71%)
897d04e — 2026-03-06 22:12:253.51 (n/a)3.49 (n/a)3.48 (n/a)3.48 (n/a)0.01 (n/a)7235.40 (n/a)7215.10 (n/a)7227.20 (n/a)7171.80 (n/a)25.51 (n/a)2395.46 (n/a)2381.12 (n/a)2377.13 (n/a)2374.41 (n/a)8.45 (n/a)

gemm_384x1536x1792_32x48x64_4cols_bcolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
e73d4ff — 2026-03-20 22:04:504.26 (-1.89%)3.82 (-8.26%)3.78 (-12.46%)3.62 (+2.64%)0.26 (-27.29%)2225.20 (-2.57%)2116.42 (+8.66%)2132.20 (+14.24%)1891.90 (+1.92%)134.52 (-28.45%)1117.39 (-1.89%)1002.29 (-8.26%)991.44 (-12.46%)950.01 (+2.64%)67.90 (-27.29%)
897d04e — 2026-03-06 22:12:254.34 (n/a)4.17 (n/a)4.32 (n/a)3.53 (n/a)0.36 (n/a)2284.00 (n/a)1947.78 (n/a)1866.50 (n/a)1856.20 (n/a)188.02 (n/a)1138.86 (n/a)1092.52 (n/a)1132.55 (n/a)925.53 (n/a)93.39 (n/a)

gemm_896x1792x640_32x64x80_8cols_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
e73d4ff — 2026-03-20 22:04:506.22 (-7.53%)5.12 (-18.90%)4.72 (-28.62%)4.40 (-15.27%)0.83 (+28.93%)1511.60 (+18.02%)1326.72 (+24.62%)1409.10 (+40.10%)1070.00 (+8.15%)204.18 (+64.87%)1920.71 (-7.53%)1580.75 (-18.90%)1458.57 (-28.62%)1359.65 (-15.27%)257.49 (+28.93%)
897d04e — 2026-03-06 22:12:256.72 (n/a)6.31 (n/a)6.61 (n/a)5.19 (n/a)0.65 (n/a)1280.80 (n/a)1064.58 (n/a)1005.80 (n/a)989.40 (n/a)123.84 (n/a)2077.17 (n/a)1949.09 (n/a)2043.44 (n/a)1604.60 (n/a)199.72 (n/a)

layer_norm_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.07 (-35.51%)0.05 (-42.32%)0.04 (-50.68%)0.04 (-32.65%)0.01 (-42.46%)222.20 (+48.43%)180.20 (+71.10%)191.90 (+102.64%)121.60 (+55.10%)42.79 (+37.58%)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)149.70 (n/a)105.32 (n/a)94.70 (n/a)78.40 (n/a)31.11 (n/a)

layer_norm_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.06 (-43.61%)0.05 (-47.13%)0.05 (-54.18%)0.04 (-37.47%)0.01 (-51.18%)223.40 (+59.91%)176.44 (+85.84%)169.80 (+118.25%)132.80 (+77.30%)40.65 (+44.02%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.11 (n/a)0.06 (n/a)0.02 (n/a)139.70 (n/a)94.94 (n/a)77.80 (n/a)74.90 (n/a)28.23 (n/a)

layer_norm_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.07 (-34.17%)0.06 (-36.82%)0.05 (-43.14%)0.04 (-28.31%)0.01 (-45.33%)190.00 (+39.50%)148.68 (+55.95%)150.20 (+75.88%)118.30 (+52.06%)26.71 (+13.01%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.10 (n/a)0.06 (n/a)0.02 (n/a)136.20 (n/a)95.34 (n/a)85.40 (n/a)77.80 (n/a)23.63 (n/a)

layer_norm_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.05 (-49.31%)0.05 (-44.86%)0.05 (-46.98%)0.03 (-8.55%)0.01 (-73.48%)235.10 (+9.35%)183.54 (+61.71%)178.30 (+88.68%)150.00 (+97.37%)31.24 (-45.53%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.08 (n/a)0.09 (n/a)0.04 (n/a)0.03 (n/a)215.00 (n/a)113.50 (n/a)94.50 (n/a)76.00 (n/a)57.35 (n/a)

layer_norm_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.07 (-36.49%)0.05 (-41.56%)0.05 (-47.46%)0.04 (-18.21%)0.01 (-52.43%)231.70 (+22.27%)172.04 (+61.09%)172.60 (+90.30%)122.80 (+57.44%)39.87 (-14.31%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.09 (n/a)0.04 (n/a)0.02 (n/a)189.50 (n/a)106.80 (n/a)90.70 (n/a)78.00 (n/a)46.53 (n/a)

layer_norm_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.07 (-29.63%)0.05 (-39.92%)0.05 (-47.70%)0.04 (-17.40%)0.01 (-40.86%)217.20 (+21.07%)176.82 (+61.24%)179.30 (+91.15%)121.40 (+42.15%)37.37 (-4.78%)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)179.40 (n/a)109.66 (n/a)93.80 (n/a)85.40 (n/a)39.25 (n/a)

layer_norm_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.07 (-29.65%)0.05 (-32.46%)0.04 (-44.33%)0.04 (-16.28%)0.01 (-38.27%)187.90 (+19.45%)164.82 (+45.83%)182.20 (+79.68%)120.60 (+42.05%)28.63 (+2.85%)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.08 (n/a)0.08 (n/a)0.05 (n/a)0.02 (n/a)157.30 (n/a)113.02 (n/a)101.40 (n/a)84.90 (n/a)27.83 (n/a)

layer_norm_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.04 (-52.43%)0.04 (-51.77%)0.03 (-60.01%)0.03 (-20.91%)0.00 (-82.32%)250.30 (+26.41%)235.40 (+93.65%)247.00 (+150.00%)203.10 (+110.25%)20.31 (-53.31%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)198.00 (n/a)121.56 (n/a)98.80 (n/a)96.60 (n/a)43.50 (n/a)

matrix_vector_mul_128x128_32_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:490.37 (+17.13%)0.30 (+24.87%)0.30 (+38.03%)0.23 (+12.47%)0.06 (+35.23%)
0a6c11c — 2025-12-04 00:39:100.32 (n/a)0.24 (n/a)0.22 (n/a)0.20 (n/a)0.04 (n/a)

matrix_vector_mul_128x128_32_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:100.26 (-1.81%)0.24 (+4.06%)0.24 (+0.52%)0.21 (+16.97%)0.02 (-48.16%)0.26 (-1.81%)0.24 (+4.06%)0.24 (+0.52%)0.20 (+16.97%)0.02 (-48.16%)
a4b6ffe — 2026-01-05 19:24:050.26 (n/a)0.23 (n/a)0.24 (n/a)0.18 (n/a)0.04 (n/a)0.26 (n/a)0.23 (n/a)0.24 (n/a)0.17 (n/a)0.04 (n/a)

matrix_vector_mul_128x128_32tsi_128tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
e73d4ff — 2026-03-20 22:04:500.19 (n/a)0.16 (n/a)0.17 (n/a)0.10 (n/a)0.04 (n/a)0.18 (n/a)0.16 (n/a)0.16 (n/a)0.10 (n/a)0.04 (n/a)

matrix_vector_mul_128x128_32tsi_128tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:250.37 (+5.44%)0.31 (-0.05%)0.34 (-1.99%)0.22 (-7.87%)0.06 (+15.13%)0.36 (+5.44%)0.31 (-0.05%)0.34 (-1.99%)0.21 (-7.87%)0.06 (+15.13%)
84d3478 — 2026-02-17 23:25:130.35 (n/a)0.31 (n/a)0.35 (n/a)0.23 (n/a)0.05 (n/a)0.35 (n/a)0.31 (n/a)0.34 (n/a)0.23 (n/a)0.05 (n/a)

matrix_vector_mul_2048x8192_1_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4913.57 (+3.47%)13.15 (+1.09%)13.10 (+0.16%)12.52 (-1.58%)0.37 (+153.19%)
0a6c11c — 2025-12-04 00:39:1013.11 (n/a)13.01 (n/a)13.07 (n/a)12.72 (n/a)0.15 (n/a)

matrix_vector_mul_2048x8192_1_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1013.43 (-0.07%)13.30 (+4.62%)13.27 (+1.29%)13.16 (+16.70%)0.12 (-86.69%)13.42 (-0.07%)13.29 (+4.62%)13.26 (+1.29%)13.16 (+16.70%)0.12 (-86.69%)
a4b6ffe — 2026-01-05 19:24:0513.44 (n/a)12.71 (n/a)13.10 (n/a)11.28 (n/a)0.87 (n/a)13.43 (n/a)12.70 (n/a)13.09 (n/a)11.27 (n/a)0.87 (n/a)

matrix_vector_mul_2048x8192_1_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4925.06 (+2.47%)23.71 (+1.81%)24.13 (+0.50%)20.90 (+5.41%)1.48 (-15.60%)
0a6c11c — 2025-12-04 00:39:1024.46 (n/a)23.29 (n/a)24.01 (n/a)19.83 (n/a)1.75 (n/a)

matrix_vector_mul_2048x8192_1_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1024.46 (-0.60%)21.07 (-8.03%)19.75 (-17.83%)18.08 (+1.18%)3.03 (+7.07%)24.44 (-0.60%)21.06 (-8.03%)19.74 (-17.83%)18.07 (+1.18%)3.03 (+7.07%)
a4b6ffe — 2026-01-05 19:24:0524.60 (n/a)22.91 (n/a)24.04 (n/a)17.87 (n/a)2.83 (n/a)24.59 (n/a)22.90 (n/a)24.02 (n/a)17.86 (n/a)2.83 (n/a)

matrix_vector_mul_2048x8192_1_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4942.05 (+1.54%)40.50 (+12.60%)40.49 (+2.36%)38.98 (+97.53%)0.97 (-88.09%)
0a6c11c — 2025-12-04 00:39:1041.41 (n/a)35.97 (n/a)39.55 (n/a)19.73 (n/a)8.15 (n/a)

matrix_vector_mul_2048x8192_1_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1042.58 (+1.51%)41.37 (+14.29%)41.25 (+4.89%)40.33 (+91.42%)0.92 (-89.18%)42.56 (+1.51%)41.34 (+14.29%)41.23 (+4.89%)40.30 (+91.42%)0.92 (-89.18%)
a4b6ffe — 2026-01-05 19:24:0541.95 (n/a)36.20 (n/a)39.33 (n/a)21.07 (n/a)8.54 (n/a)41.92 (n/a)36.17 (n/a)39.31 (n/a)21.05 (n/a)8.53 (n/a)

matrix_vector_mul_2048x8192_1_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4946.62 (+3.11%)44.82 (+0.17%)44.78 (+0.17%)42.60 (-3.80%)1.60 (+367.72%)
0a6c11c — 2025-12-04 00:39:1045.21 (n/a)44.75 (n/a)44.70 (n/a)44.29 (n/a)0.34 (n/a)

matrix_vector_mul_2048x8192_1_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1046.04 (+4.14%)45.06 (+3.47%)44.55 (+2.76%)44.30 (+2.84%)0.87 (+66.58%)46.01 (+4.14%)45.03 (+3.47%)44.52 (+2.76%)44.27 (+2.84%)0.87 (+66.58%)
a4b6ffe — 2026-01-05 19:24:0544.21 (n/a)43.55 (n/a)43.35 (n/a)43.07 (n/a)0.52 (n/a)44.18 (n/a)43.52 (n/a)43.32 (n/a)43.05 (n/a)0.52 (n/a)

matrix_vector_mul_2048x8192_1tsi_1024tso_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
e73d4ff — 2026-03-20 22:04:5024.71 (n/a)23.81 (n/a)23.96 (n/a)22.71 (n/a)0.73 (n/a)24.70 (n/a)23.80 (n/a)23.95 (n/a)22.70 (n/a)0.72 (n/a)

matrix_vector_mul_2048x8192_1tsi_1024tso_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2524.47 (-3.32%)24.16 (-3.13%)24.20 (-3.09%)23.84 (-3.04%)0.25 (-14.36%)24.46 (-3.32%)24.15 (-3.13%)24.18 (-3.09%)23.83 (-3.04%)0.25 (-14.36%)
84d3478 — 2026-02-17 23:25:1325.31 (n/a)24.94 (n/a)24.97 (n/a)24.59 (n/a)0.29 (n/a)25.30 (n/a)24.93 (n/a)24.95 (n/a)24.58 (n/a)0.29 (n/a)

matrix_vector_mul_2048x8192_1tsi_2048tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
e73d4ff — 2026-03-20 22:04:5012.80 (n/a)11.90 (n/a)11.74 (n/a)11.02 (n/a)0.69 (n/a)12.80 (n/a)11.89 (n/a)11.74 (n/a)11.02 (n/a)0.69 (n/a)

matrix_vector_mul_2048x8192_1tsi_2048tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2513.25 (-3.04%)13.02 (-2.92%)13.10 (-2.64%)12.56 (-4.14%)0.27 (+32.55%)13.24 (-3.04%)13.01 (-2.92%)13.09 (-2.64%)12.55 (-4.14%)0.27 (+32.55%)
84d3478 — 2026-02-17 23:25:1313.66 (n/a)13.41 (n/a)13.45 (n/a)13.10 (n/a)0.20 (n/a)13.65 (n/a)13.40 (n/a)13.44 (n/a)13.09 (n/a)0.20 (n/a)

matrix_vector_mul_2048x8192_1tsi_256tso_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
e73d4ff — 2026-03-20 22:04:5045.12 (n/a)42.32 (n/a)41.91 (n/a)41.16 (n/a)1.63 (n/a)45.09 (n/a)42.29 (n/a)41.89 (n/a)41.14 (n/a)1.63 (n/a)

matrix_vector_mul_2048x8192_1tsi_256tso_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2547.71 (+3.51%)45.38 (+9.55%)45.22 (-1.27%)43.47 (+75.24%)1.52 (-83.67%)47.68 (+3.51%)45.35 (+9.55%)45.19 (-1.27%)43.45 (+75.24%)1.52 (-83.67%)
84d3478 — 2026-02-17 23:25:1346.09 (n/a)41.43 (n/a)45.80 (n/a)24.81 (n/a)9.31 (n/a)46.06 (n/a)41.40 (n/a)45.77 (n/a)24.79 (n/a)9.30 (n/a)

matrix_vector_mul_2048x8192_1tsi_512tso_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
e73d4ff — 2026-03-20 22:04:5040.36 (n/a)38.06 (n/a)38.30 (n/a)36.63 (n/a)1.52 (n/a)40.34 (n/a)38.04 (n/a)38.28 (n/a)36.61 (n/a)1.52 (n/a)

matrix_vector_mul_2048x8192_1tsi_512tso_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2540.84 (-0.32%)39.53 (-1.25%)39.40 (-1.97%)38.87 (-0.69%)0.77 (-6.24%)40.81 (-0.32%)39.50 (-1.25%)39.37 (-1.97%)38.84 (-0.69%)0.77 (-6.24%)
84d3478 — 2026-02-17 23:25:1340.97 (n/a)40.03 (n/a)40.19 (n/a)39.14 (n/a)0.82 (n/a)40.95 (n/a)40.00 (n/a)40.17 (n/a)39.11 (n/a)0.82 (n/a)

matrix_vector_mul_8192x2048_4_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4913.29 (+1.63%)13.19 (+6.08%)13.19 (+1.17%)13.09 (+26.86%)0.08 (-92.94%)
0a6c11c — 2025-12-04 00:39:1013.08 (n/a)12.44 (n/a)13.03 (n/a)10.32 (n/a)1.07 (n/a)

matrix_vector_mul_8192x2048_4_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1013.68 (+3.42%)13.30 (+8.54%)13.38 (+7.25%)12.72 (+14.96%)0.35 (-66.09%)13.67 (+3.42%)13.29 (+8.54%)13.38 (+7.25%)12.71 (+14.96%)0.35 (-66.09%)
a4b6ffe — 2026-01-05 19:24:0513.23 (n/a)12.25 (n/a)12.48 (n/a)11.07 (n/a)1.04 (n/a)13.22 (n/a)12.24 (n/a)12.47 (n/a)11.06 (n/a)1.04 (n/a)

matrix_vector_mul_8192x2048_4_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4925.58 (+4.24%)24.23 (+6.59%)24.03 (+0.29%)23.58 (+37.29%)0.70 (-74.97%)
0a6c11c — 2025-12-04 00:39:1024.54 (n/a)22.73 (n/a)23.96 (n/a)17.18 (n/a)2.79 (n/a)

matrix_vector_mul_8192x2048_4_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1025.08 (+1.32%)24.68 (+13.42%)24.61 (+3.26%)24.47 (+48.32%)0.24 (-93.56%)25.07 (+1.32%)24.67 (+13.42%)24.59 (+3.26%)24.46 (+48.32%)0.24 (-93.56%)
a4b6ffe — 2026-01-05 19:24:0524.76 (n/a)21.76 (n/a)23.83 (n/a)16.50 (n/a)3.74 (n/a)24.74 (n/a)21.75 (n/a)23.82 (n/a)16.49 (n/a)3.73 (n/a)

matrix_vector_mul_8192x2048_4_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4940.94 (-2.19%)39.74 (-1.10%)39.76 (-0.60%)38.17 (-2.14%)1.04 (+4.39%)
0a6c11c — 2025-12-04 00:39:1041.86 (n/a)40.18 (n/a)40.00 (n/a)39.00 (n/a)1.00 (n/a)

matrix_vector_mul_8192x2048_4_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1040.74 (-1.09%)37.16 (-7.15%)40.25 (+1.47%)24.51 (-37.44%)7.07 (+736.13%)40.72 (-1.09%)37.13 (-7.15%)40.23 (+1.47%)24.50 (-37.44%)7.07 (+736.13%)
a4b6ffe — 2026-01-05 19:24:0541.19 (n/a)40.02 (n/a)39.67 (n/a)39.18 (n/a)0.85 (n/a)41.17 (n/a)39.99 (n/a)39.64 (n/a)39.16 (n/a)0.85 (n/a)

matrix_vector_mul_8192x2048_4_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4946.71 (-1.04%)44.91 (-2.34%)44.71 (-2.98%)43.80 (-1.44%)0.96 (+6.93%)
0a6c11c — 2025-12-04 00:39:1047.20 (n/a)45.98 (n/a)46.08 (n/a)44.44 (n/a)0.90 (n/a)

matrix_vector_mul_8192x2048_4_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1046.28 (+5.68%)44.89 (+14.59%)44.74 (+4.26%)44.20 (+62.80%)0.84 (-87.96%)46.25 (+5.68%)44.86 (+14.59%)44.71 (+4.26%)44.17 (+62.80%)0.84 (-87.96%)
a4b6ffe — 2026-01-05 19:24:0543.79 (n/a)39.17 (n/a)42.91 (n/a)27.15 (n/a)6.98 (n/a)43.77 (n/a)39.15 (n/a)42.88 (n/a)27.13 (n/a)6.98 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
e73d4ff — 2026-03-20 22:04:5013.59 (n/a)12.79 (n/a)12.71 (n/a)12.14 (n/a)0.58 (n/a)13.58 (n/a)12.78 (n/a)12.71 (n/a)12.13 (n/a)0.58 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2513.34 (-0.79%)13.20 (-0.36%)13.18 (-1.55%)13.12 (+1.75%)0.08 (-65.93%)13.33 (-0.79%)13.19 (-0.36%)13.17 (-1.55%)13.11 (+1.75%)0.08 (-65.93%)
84d3478 — 2026-02-17 23:25:1313.45 (n/a)13.25 (n/a)13.39 (n/a)12.90 (n/a)0.25 (n/a)13.44 (n/a)13.24 (n/a)13.38 (n/a)12.89 (n/a)0.25 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
e73d4ff — 2026-03-20 22:04:5024.44 (n/a)23.47 (n/a)23.14 (n/a)22.94 (n/a)0.65 (n/a)24.42 (n/a)23.45 (n/a)23.13 (n/a)22.92 (n/a)0.65 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2524.10 (-5.45%)23.07 (-1.85%)24.01 (-2.88%)19.26 (+7.47%)2.13 (-32.15%)24.08 (-5.45%)23.05 (-1.85%)23.99 (-2.88%)19.25 (+7.47%)2.13 (-32.15%)
84d3478 — 2026-02-17 23:25:1325.48 (n/a)23.50 (n/a)24.72 (n/a)17.92 (n/a)3.14 (n/a)25.47 (n/a)23.48 (n/a)24.71 (n/a)17.91 (n/a)3.14 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
e73d4ff — 2026-03-20 22:04:5041.99 (n/a)39.63 (n/a)38.58 (n/a)38.29 (n/a)1.74 (n/a)41.96 (n/a)39.60 (n/a)38.56 (n/a)38.26 (n/a)1.74 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2540.80 (-0.16%)39.58 (-1.93%)39.75 (-1.71%)38.28 (-2.50%)1.10 (+67.33%)40.78 (-0.16%)39.55 (-1.93%)39.73 (-1.71%)38.26 (-2.50%)1.10 (+67.33%)
84d3478 — 2026-02-17 23:25:1340.87 (n/a)40.36 (n/a)40.45 (n/a)39.26 (n/a)0.66 (n/a)40.84 (n/a)40.33 (n/a)40.42 (n/a)39.24 (n/a)0.65 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
e73d4ff — 2026-03-20 22:04:5044.06 (n/a)39.83 (n/a)40.91 (n/a)33.47 (n/a)4.07 (n/a)44.04 (n/a)39.81 (n/a)40.89 (n/a)33.45 (n/a)4.07 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2549.31 (+6.27%)46.12 (+1.67%)45.30 (-0.27%)44.62 (+1.87%)1.85 (+85.10%)49.28 (+6.27%)46.10 (+1.67%)45.28 (-0.27%)44.59 (+1.87%)1.85 (+85.10%)
84d3478 — 2026-02-17 23:25:1346.40 (n/a)45.37 (n/a)45.43 (n/a)43.80 (n/a)1.00 (n/a)46.37 (n/a)45.34 (n/a)45.40 (n/a)43.78 (n/a)1.00 (n/a)

mem_copy_16_cores_2_chans_2048_tile_128_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.04 (-37.26%)0.04 (-38.79%)0.04 (-38.91%)0.03 (-35.40%)0.00 (-35.20%)268.20 (+54.76%)224.72 (+62.91%)218.60 (+63.62%)196.70 (+59.40%)29.07 (+57.96%)
130b6ea — 2025-12-05 21:15:490.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)173.30 (n/a)137.94 (n/a)133.60 (n/a)123.40 (n/a)18.40 (n/a)

mem_copy_16_cores_2_chans_2048_tile_128_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-2.12%)0.06 (-16.44%)0.06 (-21.48%)0.03 (-23.84%)0.02 (+10.15%)247.60 (+31.28%)148.38 (+24.08%)128.30 (+27.41%)100.10 (+2.14%)58.70 (+50.73%)
84d3478 — 2026-02-17 23:25:130.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)188.60 (n/a)119.58 (n/a)100.70 (n/a)98.00 (n/a)38.94 (n/a)

mem_copy_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (-16.99%)0.05 (-12.47%)0.05 (-7.32%)0.04 (-11.17%)0.00 (-29.61%)197.00 (+12.57%)171.00 (+13.70%)164.00 (+7.89%)153.00 (+20.47%)17.49 (-4.32%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)175.00 (n/a)150.40 (n/a)152.00 (n/a)127.00 (n/a)18.28 (n/a)

mem_copy_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (+5.56%)0.05 (-1.59%)0.05 (-1.39%)0.03 (-14.07%)0.01 (+25.03%)263.00 (+16.37%)166.00 (+4.53%)144.00 (+1.41%)126.00 (-5.26%)49.38 (+43.17%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)226.00 (n/a)158.80 (n/a)142.00 (n/a)133.00 (n/a)34.49 (n/a)

mem_copy_1_cores_1_chans_2048_tile_2048_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.07 (-16.37%)0.06 (-18.76%)0.05 (-19.09%)0.05 (-16.54%)0.01 (+5.83%)180.00 (+19.84%)150.86 (+23.98%)155.20 (+23.57%)117.10 (+19.61%)28.71 (+54.72%)
130b6ea — 2025-12-05 21:15:490.08 (n/a)0.07 (n/a)0.07 (n/a)0.05 (n/a)0.01 (n/a)150.20 (n/a)121.68 (n/a)125.60 (n/a)97.90 (n/a)18.56 (n/a)

mem_copy_1_cores_1_chans_2048_tile_2048_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (-0.24%)0.08 (-17.53%)0.08 (-27.05%)0.06 (-15.43%)0.02 (+4.62%)148.50 (+18.23%)113.02 (+23.38%)103.60 (+37.04%)74.10 (+0.14%)31.69 (+33.33%)
84d3478 — 2026-02-17 23:25:130.11 (n/a)0.09 (n/a)0.11 (n/a)0.07 (n/a)0.02 (n/a)125.60 (n/a)91.60 (n/a)75.60 (n/a)74.00 (n/a)23.77 (n/a)

mem_copy_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (-15.32%)0.05 (-17.09%)0.05 (-23.81%)0.04 (-15.54%)0.01 (-18.18%)193.00 (+18.40%)160.80 (+20.36%)168.00 (+31.25%)124.00 (+18.10%)22.92 (+11.39%)
0999a9a — 2025-11-13 19:24:080.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)163.00 (n/a)133.60 (n/a)128.00 (n/a)105.00 (n/a)20.58 (n/a)

mem_copy_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (-23.02%)0.05 (-6.98%)0.05 (-2.13%)0.04 (-15.31%)0.01 (-33.29%)196.00 (+18.07%)147.00 (+6.52%)141.00 (+2.17%)126.00 (+29.90%)25.27 (+8.13%)
0999a9a — 2025-11-13 19:24:080.08 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)166.00 (n/a)138.00 (n/a)138.00 (n/a)97.00 (n/a)23.37 (n/a)

mem_copy_2_cores_1_chans_2048_tile_1024_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.05 (-55.54%)0.04 (-53.76%)0.04 (-49.10%)0.03 (-57.37%)0.01 (-38.51%)257.70 (+134.49%)214.38 (+118.09%)198.10 (+96.53%)178.60 (+124.94%)35.36 (+235.40%)
130b6ea — 2025-12-05 21:15:490.10 (n/a)0.08 (n/a)0.08 (n/a)0.07 (n/a)0.01 (n/a)109.90 (n/a)98.30 (n/a)100.80 (n/a)79.40 (n/a)10.54 (n/a)

mem_copy_2_cores_1_chans_2048_tile_1024_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (+27.03%)0.09 (+25.79%)0.11 (+27.09%)0.05 (+12.74%)0.03 (+43.93%)161.90 (-11.29%)94.34 (-18.35%)76.30 (-21.34%)74.00 (-21.36%)38.01 (+0.49%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)182.50 (n/a)115.54 (n/a)97.00 (n/a)94.10 (n/a)37.82 (n/a)

mem_copy_2_cores_2_chans_2048_tile_1024_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.07 (-28.49%)0.05 (-35.72%)0.05 (-38.10%)0.03 (-38.87%)0.01 (-2.18%)245.60 (+63.62%)182.46 (+59.21%)180.50 (+61.59%)117.30 (+39.81%)46.13 (+117.24%)
130b6ea — 2025-12-05 21:15:490.10 (n/a)0.07 (n/a)0.07 (n/a)0.05 (n/a)0.01 (n/a)150.10 (n/a)114.60 (n/a)111.70 (n/a)83.90 (n/a)21.23 (n/a)

mem_copy_2_cores_2_chans_2048_tile_1024_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.10 (+2.70%)0.08 (-5.38%)0.09 (+2.10%)0.04 (-43.30%)0.02 (+170.50%)201.60 (+76.38%)112.08 (+16.53%)90.80 (-2.05%)84.60 (-2.65%)50.14 (+375.75%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.09 (n/a)0.09 (n/a)0.07 (n/a)0.01 (n/a)114.30 (n/a)96.18 (n/a)92.70 (n/a)86.90 (n/a)10.54 (n/a)

mem_copy_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.07 (+11.97%)0.05 (+3.09%)0.05 (+2.10%)0.03 (-11.35%)0.01 (+41.65%)229.00 (+12.81%)156.00 (-0.51%)143.00 (-2.05%)117.00 (-10.69%)38.21 (+48.71%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)203.00 (n/a)156.80 (n/a)146.00 (n/a)131.00 (n/a)25.69 (n/a)

mem_copy_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (+9.02%)0.05 (+5.59%)0.05 (+0.00%)0.04 (+3.63%)0.01 (+25.48%)193.00 (-3.50%)147.40 (-4.66%)147.00 (+0.00%)122.00 (-8.27%)25.24 (+7.48%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)200.00 (n/a)154.60 (n/a)147.00 (n/a)133.00 (n/a)23.48 (n/a)

mem_copy_4_cores_1_chans_2048_tile_512_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.06 (-36.71%)0.04 (-39.77%)0.05 (-34.21%)0.04 (-44.36%)0.01 (-12.35%)233.20 (+79.80%)190.84 (+68.44%)177.60 (+51.92%)142.80 (+57.96%)38.35 (+155.02%)
130b6ea — 2025-12-05 21:15:490.09 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)129.70 (n/a)113.30 (n/a)116.90 (n/a)90.40 (n/a)15.04 (n/a)

mem_copy_4_cores_1_chans_2048_tile_512_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.09 (-5.87%)0.08 (-3.61%)0.08 (-0.79%)0.06 (-12.74%)0.01 (+22.04%)146.30 (+14.66%)108.02 (+4.96%)99.30 (+0.81%)95.10 (+6.26%)21.49 (+48.38%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)127.60 (n/a)102.92 (n/a)98.50 (n/a)89.50 (n/a)14.49 (n/a)

mem_copy_4_cores_2_chans_2048_tile_512_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.07 (-13.30%)0.05 (-35.24%)0.04 (-43.99%)0.03 (-49.00%)0.02 (+67.34%)263.70 (+96.06%)192.50 (+65.04%)208.60 (+78.60%)113.00 (+15.31%)57.52 (+266.09%)
130b6ea — 2025-12-05 21:15:490.08 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)134.50 (n/a)116.64 (n/a)116.80 (n/a)98.00 (n/a)15.71 (n/a)

mem_copy_4_cores_2_chans_2048_tile_512_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (+13.04%)0.09 (+17.19%)0.10 (+20.43%)0.06 (+17.23%)0.02 (+16.34%)142.80 (-14.70%)91.94 (-14.71%)79.00 (-17.02%)77.70 (-11.50%)28.50 (-14.82%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)167.40 (n/a)107.80 (n/a)95.20 (n/a)87.80 (n/a)33.46 (n/a)

mem_copy_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (-13.48%)0.05 (-17.79%)0.05 (-15.33%)0.03 (-25.09%)0.01 (+6.67%)267.00 (+33.50%)176.80 (+24.51%)150.00 (+18.11%)141.00 (+15.57%)47.41 (+61.41%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.06 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)200.00 (n/a)142.00 (n/a)127.00 (n/a)122.00 (n/a)29.37 (n/a)

mem_copy_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (+13.38%)0.05 (+23.85%)0.05 (+37.91%)0.05 (+30.72%)0.00 (-56.85%)166.00 (-23.50%)153.40 (-20.60%)153.00 (-27.49%)142.00 (-11.80%)7.63 (-70.60%)
0999a9a — 2025-11-13 19:24:080.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)217.00 (n/a)193.20 (n/a)211.00 (n/a)161.00 (n/a)25.96 (n/a)

mem_copy_8_cores_1_chans_2048_tile_256_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.05 (-47.32%)0.04 (-43.41%)0.04 (-43.65%)0.04 (-40.15%)0.00 (-73.62%)196.00 (+67.09%)189.56 (+75.84%)192.60 (+77.51%)178.60 (+89.80%)6.91 (-16.65%)
130b6ea — 2025-12-05 21:15:490.09 (n/a)0.08 (n/a)0.08 (n/a)0.07 (n/a)0.01 (n/a)117.30 (n/a)107.80 (n/a)108.50 (n/a)94.10 (n/a)8.29 (n/a)

mem_copy_8_cores_1_chans_2048_tile_256_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-0.39%)0.07 (+10.07%)0.08 (+1.07%)0.06 (+38.38%)0.01 (-47.35%)136.70 (-27.75%)110.64 (-13.07%)104.70 (-1.13%)103.00 (+0.39%)14.59 (-60.78%)
84d3478 — 2026-02-17 23:25:130.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)189.20 (n/a)127.28 (n/a)105.90 (n/a)102.60 (n/a)37.20 (n/a)

mem_copy_8_cores_2_chans_2048_tile_256_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.05 (-43.19%)0.04 (-42.83%)0.04 (-36.38%)0.02 (-59.47%)0.01 (-3.09%)336.90 (+146.81%)221.36 (+83.15%)196.70 (+57.23%)161.20 (+75.98%)69.74 (+343.56%)
130b6ea — 2025-12-05 21:15:490.09 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)136.50 (n/a)120.86 (n/a)125.10 (n/a)91.60 (n/a)15.72 (n/a)

mem_copy_8_cores_2_chans_2048_tile_256_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-17.85%)0.07 (-24.66%)0.08 (-18.40%)0.04 (-32.08%)0.02 (+21.78%)185.30 (+47.18%)127.72 (+37.87%)102.00 (+22.45%)100.40 (+21.70%)38.66 (+106.34%)
84d3478 — 2026-02-17 23:25:130.10 (n/a)0.09 (n/a)0.10 (n/a)0.07 (n/a)0.01 (n/a)125.90 (n/a)92.64 (n/a)83.30 (n/a)82.50 (n/a)18.74 (n/a)

mha

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.18 (-0.43%)0.18 (-0.55%)0.18 (-0.54%)0.18 (-0.79%)0.00 (+267.75%)47630.10 (+0.80%)47476.36 (+0.56%)47467.50 (+0.54%)47372.10 (+0.43%)106.89 (+272.38%)
130b6ea — 2025-12-05 21:15:490.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47253.90 (n/a)47213.70 (n/a)47210.90 (n/a)47167.70 (n/a)28.71 (n/a)

mha0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
84d3478 — 2026-02-17 23:25:130.18 (-0.51%)0.18 (-0.36%)0.18 (-0.41%)0.18 (-0.21%)0.00 (-35.21%)47468.60 (+0.21%)47403.94 (+0.36%)47426.00 (+0.41%)47314.40 (+0.51%)68.83 (-34.73%)
1531781 — 2026-01-23 18:49:240.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47367.50 (n/a)47232.68 (n/a)47232.00 (n/a)47073.00 (n/a)105.45 (n/a)

mha_16384_64_1_8_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47533.70 (n/a)47380.56 (n/a)47337.90 (n/a)47268.00 (n/a)110.60 (n/a)

relu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.05 (-39.97%)0.04 (-41.49%)0.05 (-47.23%)0.04 (-34.75%)0.01 (-59.09%)231.50 (+53.31%)191.58 (+65.67%)176.90 (+89.40%)153.30 (+66.63%)33.87 (+7.34%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.07 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)151.00 (n/a)115.64 (n/a)93.40 (n/a)92.00 (n/a)31.56 (n/a)

relu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.06 (-30.00%)0.04 (-46.18%)0.04 (-56.18%)0.04 (-43.83%)0.01 (+0.05%)232.70 (+78.04%)197.48 (+90.54%)215.10 (+128.34%)133.30 (+42.87%)40.61 (+152.84%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.09 (n/a)0.06 (n/a)0.01 (n/a)130.70 (n/a)103.64 (n/a)94.20 (n/a)93.30 (n/a)16.06 (n/a)

relu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.05 (-54.58%)0.04 (-54.15%)0.04 (-59.38%)0.04 (-23.51%)0.01 (-76.35%)225.80 (+30.75%)195.32 (+100.53%)192.90 (+146.36%)166.40 (+120.11%)27.92 (-33.81%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.10 (n/a)0.05 (n/a)0.03 (n/a)172.70 (n/a)97.40 (n/a)78.30 (n/a)75.60 (n/a)42.18 (n/a)

relu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.06 (-26.08%)0.04 (-39.98%)0.04 (-47.85%)0.04 (-12.25%)0.01 (-43.16%)230.80 (+13.98%)197.56 (+60.57%)198.10 (+91.77%)137.10 (+35.34%)38.13 (-14.20%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)202.50 (n/a)123.04 (n/a)103.30 (n/a)101.30 (n/a)44.44 (n/a)

rms_norm_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.07 (-14.91%)0.05 (-40.28%)0.04 (-50.33%)0.03 (-55.96%)0.02 (+46.30%)324.50 (+127.08%)201.72 (+84.66%)196.00 (+101.44%)113.00 (+17.46%)79.37 (+294.27%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)142.90 (n/a)109.24 (n/a)97.30 (n/a)96.20 (n/a)20.13 (n/a)

rms_norm_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.09 (+1.35%)0.07 (-15.92%)0.06 (-23.97%)0.04 (-25.37%)0.02 (+39.40%)196.70 (+33.99%)133.46 (+23.44%)129.20 (+31.57%)95.60 (-1.34%)39.71 (+83.00%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)146.80 (n/a)108.12 (n/a)98.20 (n/a)96.90 (n/a)21.70 (n/a)

rms_norm_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.06 (-52.32%)0.04 (-55.15%)0.04 (-57.93%)0.03 (-52.79%)0.01 (-52.56%)265.10 (+111.74%)189.92 (+122.55%)185.50 (+137.82%)142.80 (+109.69%)46.21 (+103.74%)
897d04e — 2026-03-06 22:12:250.12 (n/a)0.10 (n/a)0.10 (n/a)0.07 (n/a)0.02 (n/a)125.20 (n/a)85.34 (n/a)78.00 (n/a)68.10 (n/a)22.68 (n/a)

rms_norm_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.06 (-24.87%)0.05 (-35.29%)0.05 (-37.95%)0.04 (-39.19%)0.01 (-10.90%)230.30 (+64.50%)166.40 (+56.98%)157.10 (+61.13%)128.70 (+33.09%)38.28 (+101.28%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)140.00 (n/a)106.00 (n/a)97.50 (n/a)96.70 (n/a)19.02 (n/a)

rms_norm_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.06 (-35.08%)0.05 (-42.26%)0.05 (-43.52%)0.04 (-30.10%)0.01 (-44.57%)217.70 (+43.04%)181.98 (+71.26%)169.30 (+77.09%)143.80 (+54.13%)31.94 (+24.25%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.01 (n/a)152.20 (n/a)106.26 (n/a)95.60 (n/a)93.30 (n/a)25.71 (n/a)

rms_norm_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.05 (-43.41%)0.04 (-53.93%)0.04 (-58.18%)0.03 (-51.33%)0.01 (-31.11%)268.00 (+105.52%)226.64 (+119.02%)233.20 (+139.18%)169.00 (+76.59%)36.04 (+138.99%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)130.40 (n/a)103.48 (n/a)97.50 (n/a)95.70 (n/a)15.08 (n/a)

rms_norm_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.04 (-49.76%)0.03 (-52.64%)0.04 (-51.71%)0.02 (-55.30%)0.01 (-46.24%)425.80 (+123.75%)256.68 (+114.90%)215.10 (+107.23%)189.50 (+99.05%)96.24 (+141.85%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)190.30 (n/a)119.44 (n/a)103.80 (n/a)95.20 (n/a)39.79 (n/a)

rms_norm_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.04 (-62.57%)0.03 (-62.33%)0.04 (-63.41%)0.02 (-52.70%)0.01 (-70.41%)416.30 (+111.43%)262.08 (+149.84%)227.00 (+173.16%)207.90 (+167.22%)86.75 (+68.49%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.10 (n/a)0.04 (n/a)0.03 (n/a)196.90 (n/a)104.90 (n/a)83.10 (n/a)77.80 (n/a)51.49 (n/a)

rope_1_cols_2_channels_4096_tile_4096_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.19 (-21.66%)0.17 (-8.53%)0.18 (+2.09%)0.14 (-6.71%)0.02 (-38.07%)174.10 (+7.20%)150.06 (+8.11%)139.60 (-2.04%)129.80 (+27.63%)20.36 (-10.65%)
d48746f — 2026-01-12 22:17:100.24 (n/a)0.18 (n/a)0.17 (n/a)0.15 (n/a)0.04 (n/a)162.40 (n/a)138.80 (n/a)142.50 (n/a)101.70 (n/a)22.79 (n/a)

rope_1c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.86 (-18.58%)0.67 (-24.30%)0.65 (-28.17%)0.39 (-45.33%)0.20 (+34.97%)253.00 (+82.94%)159.58 (+40.55%)150.40 (+39.13%)113.70 (+22.79%)56.75 (+195.87%)
897d04e — 2026-03-06 22:12:251.06 (n/a)0.89 (n/a)0.91 (n/a)0.71 (n/a)0.15 (n/a)138.30 (n/a)113.54 (n/a)108.10 (n/a)92.60 (n/a)19.18 (n/a)

rope_1c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.56 (-8.58%)0.42 (-16.92%)0.39 (-19.15%)0.26 (-40.64%)0.12 (+88.15%)279.20 (+68.50%)190.32 (+28.23%)188.40 (+23.70%)130.90 (+9.36%)59.76 (+249.42%)
897d04e — 2026-03-06 22:12:250.62 (n/a)0.50 (n/a)0.48 (n/a)0.44 (n/a)0.07 (n/a)165.70 (n/a)148.42 (n/a)152.30 (n/a)119.70 (n/a)17.10 (n/a)

rope_2_cols_2_channels_4096_tile_2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.21 (-0.56%)0.17 (-10.62%)0.16 (-13.58%)0.13 (-13.33%)0.03 (+17.05%)183.70 (+15.39%)151.76 (+12.77%)155.20 (+15.73%)117.30 (+0.60%)24.24 (+35.73%)
d48746f — 2026-01-12 22:17:100.21 (n/a)0.19 (n/a)0.18 (n/a)0.15 (n/a)0.02 (n/a)159.20 (n/a)134.58 (n/a)134.10 (n/a)116.60 (n/a)17.86 (n/a)

rope_2c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.76 (-29.72%)0.61 (-32.14%)0.52 (-44.32%)0.50 (-12.42%)0.14 (-31.45%)196.90 (+14.14%)168.20 (+45.45%)190.30 (+79.70%)129.50 (+42.31%)35.41 (+7.01%)
897d04e — 2026-03-06 22:12:251.08 (n/a)0.90 (n/a)0.93 (n/a)0.57 (n/a)0.20 (n/a)172.50 (n/a)115.64 (n/a)105.90 (n/a)91.00 (n/a)33.09 (n/a)

rope_2c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.58 (-33.73%)0.45 (-27.80%)0.47 (-9.25%)0.32 (-34.35%)0.10 (-43.30%)230.80 (+52.34%)169.24 (+36.42%)158.30 (+10.16%)127.10 (+50.95%)40.33 (+30.25%)
897d04e — 2026-03-06 22:12:250.88 (n/a)0.63 (n/a)0.51 (n/a)0.49 (n/a)0.18 (n/a)151.50 (n/a)124.06 (n/a)143.70 (n/a)84.20 (n/a)30.96 (n/a)

rope_4_cols_2_channels_4096_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.22 (+18.92%)0.17 (+6.24%)0.17 (+5.57%)0.14 (-0.50%)0.03 (+90.77%)171.10 (+0.47%)147.40 (-4.39%)148.10 (-5.31%)110.10 (-15.89%)23.59 (+60.67%)
d48746f — 2026-01-12 22:17:100.19 (n/a)0.16 (n/a)0.16 (n/a)0.14 (n/a)0.02 (n/a)170.30 (n/a)154.16 (n/a)156.40 (n/a)130.90 (n/a)14.69 (n/a)

rope_8_cols_2_channels_4096_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.17 (-1.30%)0.15 (+6.63%)0.15 (-8.78%)0.14 (+39.68%)0.01 (-67.35%)174.90 (-28.41%)163.42 (-10.36%)167.40 (+9.63%)146.50 (+1.31%)11.19 (-76.17%)
d48746f — 2026-01-12 22:17:100.17 (n/a)0.14 (n/a)0.16 (n/a)0.10 (n/a)0.03 (n/a)244.30 (n/a)182.30 (n/a)152.70 (n/a)144.60 (n/a)46.98 (n/a)

rope_8c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.58 (-45.65%)0.45 (-49.22%)0.43 (-57.85%)0.39 (-27.38%)0.08 (-66.62%)251.80 (+37.75%)223.52 (+87.55%)227.50 (+137.23%)169.80 (+83.97%)33.13 (-14.59%)
897d04e — 2026-03-06 22:12:251.07 (n/a)0.88 (n/a)1.03 (n/a)0.54 (n/a)0.23 (n/a)182.80 (n/a)119.18 (n/a)95.90 (n/a)92.30 (n/a)38.79 (n/a)

rope_8c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.57 (+0.89%)0.43 (-10.91%)0.41 (-22.19%)0.37 (+1.11%)0.08 (-6.32%)200.70 (-1.13%)176.94 (+11.78%)180.50 (+28.56%)130.20 (-0.84%)28.59 (-8.87%)
897d04e — 2026-03-06 22:12:250.56 (n/a)0.48 (n/a)0.52 (n/a)0.36 (n/a)0.09 (n/a)203.00 (n/a)158.30 (n/a)140.40 (n/a)131.30 (n/a)31.38 (n/a)

sigmoid_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.06 (-34.83%)0.05 (-31.95%)0.05 (-37.62%)0.04 (-24.44%)0.01 (-61.57%)200.30 (+32.39%)162.56 (+42.57%)150.90 (+60.36%)143.50 (+53.48%)22.85 (-18.54%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)151.30 (n/a)114.02 (n/a)94.10 (n/a)93.50 (n/a)28.04 (n/a)

sigmoid_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.06 (-44.16%)0.05 (-42.39%)0.05 (-54.56%)0.04 (-9.05%)0.01 (-77.69%)194.10 (+9.97%)167.02 (+56.39%)169.30 (+120.16%)136.90 (+79.19%)20.66 (-53.90%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.11 (n/a)0.05 (n/a)0.03 (n/a)176.50 (n/a)106.80 (n/a)76.90 (n/a)76.40 (n/a)44.81 (n/a)

sigmoid_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.05 (-42.08%)0.04 (-44.51%)0.05 (-41.79%)0.03 (-40.22%)0.01 (-39.36%)287.10 (+67.31%)203.60 (+80.30%)172.10 (+71.76%)155.90 (+72.65%)55.64 (+67.13%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.08 (n/a)0.05 (n/a)0.02 (n/a)171.60 (n/a)112.92 (n/a)100.20 (n/a)90.30 (n/a)33.29 (n/a)

sigmoid_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.07 (-26.25%)0.05 (-41.44%)0.05 (-53.83%)0.03 (-10.17%)0.01 (-45.56%)237.20 (+11.31%)172.28 (+57.79%)180.70 (+116.67%)112.10 (+35.71%)46.50 (-19.96%)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.09 (n/a)0.10 (n/a)0.04 (n/a)0.03 (n/a)213.10 (n/a)109.18 (n/a)83.40 (n/a)82.60 (n/a)58.09 (n/a)

silu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.05 (-55.00%)0.04 (-48.36%)0.04 (-60.07%)0.04 (-4.68%)0.00 (-89.87%)197.90 (+4.93%)190.48 (+68.87%)196.90 (+150.51%)168.30 (+122.32%)12.63 (-75.25%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.08 (n/a)0.10 (n/a)0.04 (n/a)0.03 (n/a)188.60 (n/a)112.80 (n/a)78.60 (n/a)75.70 (n/a)51.04 (n/a)

silu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.07 (-21.11%)0.04 (-41.17%)0.04 (-49.01%)0.03 (-45.49%)0.02 (+9.36%)264.90 (+83.45%)201.48 (+79.92%)213.40 (+96.14%)110.40 (+26.75%)59.28 (+146.36%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.02 (n/a)144.40 (n/a)111.98 (n/a)108.80 (n/a)87.10 (n/a)24.06 (n/a)

silu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.08 (-16.49%)0.05 (-39.49%)0.04 (-52.47%)0.03 (-29.13%)0.02 (+3.38%)234.10 (+41.11%)188.20 (+70.07%)208.40 (+110.51%)107.50 (+19.71%)49.43 (+58.70%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.08 (n/a)0.05 (n/a)0.02 (n/a)165.90 (n/a)110.66 (n/a)99.00 (n/a)89.80 (n/a)31.15 (n/a)

silu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.04 (-56.86%)0.04 (-56.58%)0.03 (-63.34%)0.03 (-26.54%)0.00 (-79.45%)248.90 (+36.16%)222.76 (+112.31%)237.90 (+172.82%)188.80 (+131.66%)27.15 (-37.77%)
897d04e — 2026-03-06 22:12:250.10 (n/a)0.09 (n/a)0.09 (n/a)0.04 (n/a)0.02 (n/a)182.80 (n/a)104.92 (n/a)87.20 (n/a)81.50 (n/a)43.64 (n/a)

softmax_1_cols_2_channels_4096_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.19 (-8.94%)0.17 (-8.66%)0.17 (-19.83%)0.14 (+4.32%)0.02 (-35.13%)115.80 (-4.14%)99.26 (+7.47%)98.60 (+24.65%)85.00 (+9.82%)13.65 (-30.86%)
84d3478 — 2026-02-17 23:25:130.21 (n/a)0.18 (n/a)0.21 (n/a)0.14 (n/a)0.04 (n/a)120.80 (n/a)92.36 (n/a)79.10 (n/a)77.40 (n/a)19.75 (n/a)

softmax_2_cols_2_channels_32768_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:501.02 (n/a)0.83 (n/a)0.89 (n/a)0.61 (n/a)0.18 (n/a)213.60 (n/a)164.16 (n/a)147.30 (n/a)128.30 (n/a)37.64 (n/a)

softmax_2_cols_2_channels_32768_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:501.17 (n/a)0.91 (n/a)0.90 (n/a)0.54 (n/a)0.25 (n/a)243.90 (n/a)155.98 (n/a)145.00 (n/a)111.80 (n/a)52.66 (n/a)

softmax_2_cols_2_channels_4096_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.20 (-0.87%)0.16 (+6.57%)0.17 (-9.55%)0.11 (+29.81%)0.03 (-44.34%)145.70 (-22.99%)104.68 (-15.00%)95.00 (+10.59%)83.80 (+0.84%)24.16 (-54.23%)
84d3478 — 2026-02-17 23:25:130.20 (n/a)0.15 (n/a)0.19 (n/a)0.09 (n/a)0.06 (n/a)189.20 (n/a)123.16 (n/a)85.90 (n/a)83.10 (n/a)52.80 (n/a)

softmax_2_cols_2_channels_4096_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.25 (+23.53%)0.18 (+9.42%)0.18 (-3.86%)0.11 (+2.99%)0.05 (+19.24%)144.10 (-2.90%)98.98 (-8.18%)92.10 (+4.07%)64.30 (-19.02%)29.58 (-7.13%)
84d3478 — 2026-02-17 23:25:130.21 (n/a)0.16 (n/a)0.19 (n/a)0.11 (n/a)0.04 (n/a)148.40 (n/a)107.80 (n/a)88.50 (n/a)79.40 (n/a)31.85 (n/a)

softmax_4_cols_4_channels_32768_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.62 (n/a)0.51 (n/a)0.53 (n/a)0.39 (n/a)0.09 (n/a)336.40 (n/a)262.78 (n/a)247.50 (n/a)212.90 (n/a)50.58 (n/a)

swiglu

No metrics available.

swiglu_decode_1x2048x2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:50n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
130b6ea — 2025-12-05 21:15:490.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)5187.20 (n/a)4720.84 (n/a)4750.90 (n/a)4015.10 (n/a)390.23 (n/a)

swiglu_decode_1x2048x2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.00 (-0.21%)0.00 (+7.74%)0.00 (+0.37%)0.00 (+8.84%)0.00 (-19.58%)5170.80 (-8.12%)3761.80 (-8.70%)3408.60 (-0.36%)3407.30 (+0.21%)787.66 (-22.98%)
84d3478 — 2026-02-17 23:25:130.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)5627.70 (n/a)4120.44 (n/a)3421.00 (n/a)3400.30 (n/a)1022.61 (n/a)

swiglu_prefill_256x2048x2048

No metrics available.

tanh_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.06 (-24.84%)0.05 (-38.33%)0.05 (-41.02%)0.04 (-39.47%)0.01 (+6.95%)214.30 (+65.23%)175.70 (+64.95%)178.10 (+69.62%)126.80 (+33.05%)31.53 (+126.00%)
897d04e — 2026-03-06 22:12:250.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)129.70 (n/a)106.52 (n/a)105.00 (n/a)95.30 (n/a)13.95 (n/a)

tanh_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.06 (-42.42%)0.05 (-47.35%)0.05 (-53.99%)0.04 (-28.57%)0.01 (-60.69%)223.90 (+40.03%)171.10 (+80.26%)165.50 (+117.48%)129.70 (+73.63%)34.83 (-5.16%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.09 (n/a)0.11 (n/a)0.05 (n/a)0.02 (n/a)159.90 (n/a)94.92 (n/a)76.10 (n/a)74.70 (n/a)36.73 (n/a)

tanh_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.07 (-16.21%)0.05 (-31.19%)0.04 (-44.38%)0.04 (-40.68%)0.02 (+86.66%)217.40 (+68.53%)173.48 (+54.07%)191.70 (+79.83%)120.90 (+19.35%)47.94 (+269.64%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.07 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)129.00 (n/a)112.60 (n/a)106.60 (n/a)101.30 (n/a)12.97 (n/a)

tanh_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.07 (-16.83%)0.04 (-43.47%)0.04 (-52.38%)0.03 (-50.11%)0.01 (+66.71%)268.00 (+100.45%)206.98 (+88.16%)214.10 (+110.11%)119.40 (+20.24%)54.29 (+274.70%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)133.70 (n/a)110.00 (n/a)101.90 (n/a)99.30 (n/a)14.49 (n/a)

transpose_2048_M_64_N_1_cols_1_channels_64_m_64_n_8_s

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:503.00 (-23.49%)2.52 (-31.52%)2.80 (-23.30%)1.93 (-44.20%)0.53 (+247.00%)272.00 (+79.18%)216.18 (+51.49%)186.90 (+30.33%)174.50 (+30.71%)48.82 (+727.73%)
130b6ea — 2025-12-05 21:15:493.93 (n/a)3.68 (n/a)3.66 (n/a)3.45 (n/a)0.15 (n/a)151.80 (n/a)142.70 (n/a)143.40 (n/a)133.50 (n/a)5.90 (n/a)

transpose_2048_M_64_N_1_cols_1_channels_64_m_64_n_8_s0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:254.43 (+0.85%)3.73 (+3.83%)4.02 (-1.30%)2.97 (+16.77%)0.65 (-24.99%)176.70 (-14.35%)144.46 (-6.24%)130.40 (+1.32%)118.40 (-0.84%)26.60 (-35.16%)
84d3478 — 2026-02-17 23:25:134.39 (n/a)3.59 (n/a)4.07 (n/a)2.54 (n/a)0.87 (n/a)206.30 (n/a)154.08 (n/a)128.70 (n/a)119.40 (n/a)41.02 (n/a)

transpose_2048_M_64_N_1_cols_2_channels_64_m_64_n_8_s

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:503.84 (-24.75%)3.35 (-13.05%)3.59 (+5.65%)2.24 (-31.90%)0.64 (-9.51%)234.40 (+46.87%)162.66 (+16.04%)145.90 (-5.38%)136.40 (+32.94%)40.64 (+79.05%)
130b6ea — 2025-12-05 21:15:495.11 (n/a)3.85 (n/a)3.40 (n/a)3.29 (n/a)0.71 (n/a)159.60 (n/a)140.18 (n/a)154.20 (n/a)102.60 (n/a)22.70 (n/a)

transpose_2048_M_64_N_1_cols_2_channels_64_m_64_n_8_s0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:255.27 (-14.18%)4.35 (-2.49%)4.45 (+2.15%)2.78 (-22.18%)0.98 (-3.38%)188.60 (+28.47%)126.98 (+4.20%)117.90 (-2.08%)99.40 (+16.53%)36.02 (+50.15%)
84d3478 — 2026-02-17 23:25:136.14 (n/a)4.46 (n/a)4.35 (n/a)3.57 (n/a)1.02 (n/a)146.80 (n/a)121.86 (n/a)120.40 (n/a)85.30 (n/a)23.99 (n/a)

weighted_rms_norm_1_cols_2_channels_2048_weights_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.09 (+12.38%)0.07 (-3.23%)0.07 (-8.92%)0.06 (+17.45%)0.01 (+3.93%)216.20 (+27.70%)186.50 (+53.90%)186.00 (+64.75%)133.30 (+33.43%)32.93 (+15.52%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.07 (n/a)0.07 (n/a)0.05 (n/a)0.01 (n/a)169.30 (n/a)121.18 (n/a)112.90 (n/a)99.90 (n/a)28.51 (n/a)

weighted_rms_norm_2_cols_2_channels_2048_weights_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.08 (-27.16%)0.05 (-45.59%)0.05 (-53.29%)0.04 (-30.73%)0.01 (-29.97%)253.00 (+80.46%)207.74 (+128.79%)210.80 (+167.51%)133.80 (+71.76%)46.03 (+66.66%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.10 (n/a)0.10 (n/a)0.06 (n/a)0.02 (n/a)140.20 (n/a)90.80 (n/a)78.80 (n/a)77.90 (n/a)27.62 (n/a)

weighted_rms_norm_4_cols_2_channels_2048_weights_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.05 (-57.24%)0.04 (-55.18%)0.04 (-57.37%)0.04 (-41.90%)0.00 (-82.39%)245.30 (+93.61%)217.16 (+143.51%)210.30 (+163.86%)203.10 (+163.08%)16.63 (-21.18%)
897d04e — 2026-03-06 22:12:250.11 (n/a)0.10 (n/a)0.10 (n/a)0.06 (n/a)0.02 (n/a)126.70 (n/a)89.18 (n/a)79.70 (n/a)77.20 (n/a)21.10 (n/a)

weighted_rms_norm_8_cols_2_channels_2048_weights_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
e73d4ff — 2026-03-20 22:04:500.05 (-36.50%)0.04 (-45.66%)0.04 (-49.92%)0.03 (-35.64%)0.01 (-44.43%)293.50 (+65.07%)230.78 (+93.25%)220.50 (+112.02%)172.10 (+67.25%)46.91 (+43.36%)
897d04e — 2026-03-06 22:12:250.08 (n/a)0.07 (n/a)0.08 (n/a)0.05 (n/a)0.01 (n/a)177.80 (n/a)119.42 (n/a)104.00 (n/a)102.90 (n/a)32.72 (n/a)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants