RMSNorm 32768 x 1024, time and energy

Recorded result

Pass. Best candidate 1.124 ms and 76.7 mJ a call, against 3.911 ms and 276.0 mJ for eager PyTorch

Recorded 2026-09-30

Execution conditions

• Energy read from the NVML counter around each window
• Triton 3.6.0, torch 2.10.0, CUDA 12.8, 70 W power limit
• Every candidate checked for correctness before timing

Historical result; a new workload requires compatibility checking and fresh Verify.

Use this as your starting point

Save the reference, describe your workload and set the metric that matters. Your project keeps the selection and its recorded conditions together.

A historical result is not a compatibility guarantee. Your configuration needs a fresh Verify run.

Evidence summary imported from the local registry on October 2, 2026. A summary is not a newly issued proof pack.