GEMM n4096 k4096
General matrix multiply (GEMM) C = A @ B.T. Captured from Llama 3.1 8B attn.o_proj.
Reported evidence · last observed 2026-08-03. Reported by source; not independently reproduced. No public source. License unknown.
Current records
Workloadsuite of 43 cases · mean latency86 cases
Not measured on H100 for this workload. Challenges →
114.3µs56% SOL1.00×Reported · license unknown · no source2026-08-03
Measured exactly what you asked. Reported by source; not independently reproduced.
215.0µs54% SOL1.05×Reported · license unknown · no source2026-08-17
Measured exactly what you asked. Reported by source; not independently reproduced.
315.2µs54% SOL1.06×Reported · license unknown · no source2026-08-03
Measured exactly what you asked. Reported by source; not independently reproduced.
415.2µs53% SOL1.07×Reported · license unknown · no source2026-07-31
Measured exactly what you asked. Reported by source; not independently reproduced.
515.2µs54% SOL1.07×Reported · license unknown · no source2026-07-29
Measured exactly what you asked. Reported by source; not independently reproduced.
615.3µs53% SOL1.07×Reported · license unknown · no source2026-07-28
Measured exactly what you asked. Reported by source; not independently reproduced.
715.3µs53% SOL1.07×Reported · license unknown · no source2026-07-30
Measured exactly what you asked. Reported by source; not independently reproduced.
815.4µs53% SOL1.08×Reported · license unknown · no source2026-08-23
Measured exactly what you asked. Reported by source; not independently reproduced.
915.5µs52% SOL1.09×Reported · license unknown · no source2026-07-27
Measured exactly what you asked. Reported by source; not independently reproduced.
1015.5µs52% SOL1.09×Reported · license unknown · no source2026-08-24
Measured exactly what you asked. Reported by source; not independently reproduced.
10=15.5µs52% SOL1.09×Reported · license unknown · no source2026-08-20
Measured exactly what you asked; tied ranks share a number. Reported by source; not independently reproduced.
10=15.5µs52% SOL1.09×Reported · license unknown · no source2026-08-05
Measured exactly what you asked; tied ranks share a number. Reported by source; not independently reproduced.
1115.6µs52% SOL1.09×Reported · license unknown · no source2026-08-03
Measured exactly what you asked. Reported by source; not independently reproduced.
1215.6µs52% SOL1.09×Reported · license unknown · no source2026-07-28
Measured exactly what you asked. Reported by source; not independently reproduced.
1315.6µs52% SOL1.09×Reported · license unknown · no source2026-07-10
Measured exactly what you asked. Reported by source; not independently reproduced.
1415.7µs52% SOL1.10×Reported · license unknown · no source2026-07-30
Measured exactly what you asked. Reported by source; not independently reproduced.
1515.7µs52% SOL1.10×Reported · license unknown · no source2026-07-29
Measured exactly what you asked. Reported by source; not independently reproduced.
1615.9µs51% SOL1.11×Reported · license unknown · no source2026-07-10
Measured exactly what you asked. Reported by source; not independently reproduced.
1715.9µs51% SOL1.11×Reported · license unknown · no source2026-07-28
Measured exactly what you asked. Reported by source; not independently reproduced.
1815.9µs51% SOL1.12×Reported · license unknown · no source2026-07-18
Measured exactly what you asked. Reported by source; not independently reproduced.
1916.0µs51% SOL1.12×Reported · license unknown · no source2026-07-30
Measured exactly what you asked. Reported by source; not independently reproduced.
2016.1µs51% SOL1.13×Reported · license unknown · no source2026-08-13
Measured exactly what you asked. Reported by source; not independently reproduced.
2116.1µs51% SOL1.13×Reported · license unknown · no source2026-07-25
Measured exactly what you asked. Reported by source; not independently reproduced.
2216.3µs50% SOL1.14×Reported · license unknown · no source2026-07-30
Measured exactly what you asked. Reported by source; not independently reproduced.
2316.4µs50% SOL1.15×Reported · license unknown · no source2026-07-25
Measured exactly what you asked. Reported by source; not independently reproduced.
2416.4µs50% SOL1.15×Reported · license unknown · no source2026-07-16
Measured exactly what you asked. Reported by source; not independently reproduced.
2516.5µs50% SOL1.15×Reported · license unknown · no source2026-07-23
Measured exactly what you asked. Reported by source; not independently reproduced.
2616.5µs50% SOL1.15×Reported · license unknown · no source2026-07-21
Measured exactly what you asked. Reported by source; not independently reproduced.
2716.5µs50% SOL1.16×Reported · license unknown · no source2026-07-10
Measured exactly what you asked. Reported by source; not independently reproduced.
2816.5µs50% SOL1.16×Reported · license unknown · no source2026-07-23
Measured exactly what you asked. Reported by source; not independently reproduced.
2916.7µs49% SOL1.17×Reported · license unknown · no source2026-07-10
Measured exactly what you asked. Reported by source; not independently reproduced.
3016.7µs50% SOL1.17×Reported · license unknown · no source2026-07-28
Measured exactly what you asked. Reported by source; not independently reproduced.
3117.1µs48% SOL1.20×Reported · license unknown · no source2026-07-10
Measured exactly what you asked. Reported by source; not independently reproduced.
3217.1µs48% SOL1.20×Reported · license unknown · no source2026-08-13
Measured exactly what you asked. Reported by source; not independently reproduced.
3317.2µs48% SOL1.20×Reported · license unknown · no source2026-08-13
Measured exactly what you asked. Reported by source; not independently reproduced.
3417.2µs48% SOL1.20×Reported · license unknown · no source2026-07-19
Measured exactly what you asked. Reported by source; not independently reproduced.
3517.3µs48% SOL1.21×Reported · license unknown · no source2026-07-19
Measured exactly what you asked. Reported by source; not independently reproduced.
3617.4µs48% SOL1.22×Reported · license unknown · no source2026-07-19
Measured exactly what you asked. Reported by source; not independently reproduced.
3717.4µs48% SOL1.22×Reported · license unknown · no source2026-07-15
Measured exactly what you asked. Reported by source; not independently reproduced.
3817.4µs48% SOL1.22×Reported · license unknown · no source2026-07-19
Measured exactly what you asked. Reported by source; not independently reproduced.
3917.4µs48% SOL1.22×Reported · license unknown · no source2026-07-10
Measured exactly what you asked. Reported by source; not independently reproduced.
4017.6µs47% SOL1.23×Reported · license unknown · no source2026-07-21
Measured exactly what you asked. Reported by source; not independently reproduced.
4117.6µs47% SOL1.23×Reported · license unknown · no source2026-07-19
Measured exactly what you asked. Reported by source; not independently reproduced.
4218.0µs46% SOL1.26×Reported · license unknown · no source2026-07-25
Measured exactly what you asked. Reported by source; not independently reproduced.
4318.0µs47% SOL1.26×Reported · license unknown · no source2026-07-19
Measured exactly what you asked. Reported by source; not independently reproduced.
4418.1µs46% SOL1.27×Reported · license unknown · no source2026-07-21
Measured exactly what you asked. Reported by source; not independently reproduced.
4518.1µs46% SOL1.27×Reported · license unknown · no source2026-07-21
Measured exactly what you asked. Reported by source; not independently reproduced.
4618.1µs46% SOL1.27×Reported · license unknown · no source2026-07-15
Measured exactly what you asked. Reported by source; not independently reproduced.
4718.5µs45% SOL1.30×Reported · license unknown · no source2026-07-25
Measured exactly what you asked. Reported by source; not independently reproduced.
4818.8µs45% SOL1.32×Reported · license unknown · no source2026-07-19
Measured exactly what you asked. Reported by source; not independently reproduced.
4919.4µs45% SOL1.36×Reported · license unknown · no source2026-07-10
Measured exactly what you asked. Reported by source; not independently reproduced.
Implementations
Show all 84 implementations ›Showing all 84 implementations ⌄
Semantics
Also indexed as gemm-n4096-k4096 · reviewed as the same computation; every definition's cohorts are shown here, ranked separately.