Compare runs
cohort sha256:17f3c00416720486… · source-native profile
ranking-v1
no ranks: not comparable
Comparison identity
operation
kernelbench-l1-46-average-pooling-3d
workload
depth = 128 · width = 256 · height = 128 · channels = 32 · batch_size = 16 · fp32 · b75ddf8c
protocol
KernelBench timing scripts · 547725b0
environment
NVIDIA H100 · 2750ae38
correctness policy
52ff5e65
metric
latency mean (ns)
Context · 10 fields›
architecture
sm_90
CUDA
unknown
driver
unknown
framework
unknown
samples
100
evidence
reported
license
MIT
source
available
status
passed
observed
2026-03-05
No winner: these runs didn't measure the same thing. The rows above show what differs. Why comparable?
Add run ›
Export ›
Markdown CSV JSON
PyTorch eager · sha256:726f1d5c818f0… · observed 2026-03-05