Skip to content
KernelIndex
Search⌘K

PyTorch

python · BSD-3-Clause
holds 7 records

Use it

Source available upstream · BSD-3-ClauseView source →

No install recipe recorded for this revision.

interfacepython
Compatibility
measured onNVIDIA H100
declared hardwaredeclared only
architectures—
dtypes

Benchmark evidence

8 measurements across 1 GPU, fastest first.

Operation / workload
Hardware
Latency
Rank
Observed
Fused linear + JSDbf16 · [1024, 4096]
NVIDIA H100
9.56ms
#1 of 2
2024-10-09
Fused linear + JSDbf16 · [2048, 4096]
NVIDIA H100
18.7ms
#1 of 2
2024-10-09
Fused linear + JSDbf16 · [1024, 4096]
NVIDIA H100
26.0ms
#1 of 2
2024-10-09
Fused linear + JSDbf16 · [4096, 4096]
NVIDIA H100
37.8ms
#1 of 2
2024-10-09
Fused linear + JSDbf16 · [2048, 4096]
NVIDIA H100
49.5ms
#1 of 2
2024-10-09
Fused linear + JSDbf16 · [8192, 4096]
NVIDIA H100
75.2ms
#1 of 2
2024-10-09
Fused linear + JSDbf16 · [4096, 4096]
NVIDIA H100
98.3ms
#1 of 2
2024-10-09
Fused linear + JSDbf16 · [8192, 4096]
NVIDIA H100
195.3ms
#2 of 2
2024-10-09

Reported · How evidence levels are derived →

Source and license

sourceavailable
revision digestsha256:e2560a134df774ba570c6e62110a7de637b91c5ababdacf09bebb0e7b49ca688
license declaredBSD-3-Clause
license concludedBSD-3-Clause
imported2026-08-20

Best evidence level for this revision: reported

JSON