Skip to content
KernelIndex
Search⌘K

PyTorch

python · BSD-3-Clause
holds 9 records

Use it

Source available upstream · BSD-3-ClauseView source →

No install recipe recorded for this revision.

interfacepython
Compatibility
measured onNVIDIA B200, NVIDIA H100
declared hardwaredeclared only
architectures—
dtypes

Benchmark evidence

24 measurements across 2 GPUs, fastest first.

Operation / workload
Hardware
Latency
Rank
Observed
Jensen-Shannon divergence lossfp32 · [1024, 128256]
NVIDIA B200
2.29ms
#1 of 2
2026-05-27
Jensen-Shannon divergence lossfp32 · [8192, 4096]
NVIDIA H100
2.47ms
#2 of 2
2024-10-02
Jensen-Shannon divergence lossfp32 · [1024, 128256]
NVIDIA B200
3.79ms
#2 of 2
2026-05-27
Jensen-Shannon divergence lossfp32 · [8192, 8192]
NVIDIA H100
4.85ms
#2 of 2
2024-10-02
Jensen-Shannon divergence lossfp32 · [2048, 128256]
NVIDIA B200
4.97ms
#1 of 2
2026-05-27
Jensen-Shannon divergence lossfp32 · [8192, 4096]
NVIDIA H100
5.40ms
#2 of 2
2024-10-02
Jensen-Shannon divergence lossfp32 · [1024, 128256]
NVIDIA B200
5.92ms
#1 of 2
2026-05-27
Jensen-Shannon divergence lossfp32 · [2048, 128256]
NVIDIA B200
7.67ms
#2 of 2
2026-05-27
Jensen-Shannon divergence lossfp32 · [8192, 16384]
NVIDIA H100
9.53ms
#1 of 2
2024-10-02
Jensen-Shannon divergence lossfp32 · [4096, 128256]
NVIDIA B200
9.91ms
#1 of 2
2026-05-27
Show all 24 measurements ›
Jensen-Shannon divergence lossfp32 · [8192, 8192]
NVIDIA H100
10.6ms
#2 of 2
2024-10-02
Jensen-Shannon divergence lossfp32 · [2048, 128256]
NVIDIA B200
12.2ms
#2 of 2
2026-05-27
Jensen-Shannon divergence lossfp32 · [4096, 128256]
NVIDIA B200
15.2ms
#2 of 2
2026-05-27
Jensen-Shannon divergence lossfp32 · [8192, 32768]
NVIDIA H100
18.9ms
#1 of 2
2024-10-02
Jensen-Shannon divergence lossfp32 · [8192, 128256]
NVIDIA B200
20.0ms
#1 of 2
2026-05-27
Jensen-Shannon divergence lossfp32 · [8192, 16384]
NVIDIA H100
20.9ms
#2 of 2
2024-10-02
Jensen-Shannon divergence lossfp32 · [4096, 128256]
NVIDIA B200
24.1ms
#2 of 2
2026-05-27
Jensen-Shannon divergence lossfp32 · [8192, 128256]
NVIDIA B200
30.3ms
#2 of 2
2026-05-27
Jensen-Shannon divergence lossfp32 · [8192, 65536]
NVIDIA H100
37.7ms
#1 of 2
2024-10-02
Jensen-Shannon divergence lossfp32 · [8192, 32768]
NVIDIA H100
42.1ms
#2 of 2
2024-10-02
Jensen-Shannon divergence lossfp32 · [8192, 128256]
NVIDIA B200
50.5ms
#2 of 2
2026-05-27
Jensen-Shannon divergence lossfp32 · [8192, 131072]
NVIDIA H100
75.4ms
#1 of 2
2024-10-02
Jensen-Shannon divergence lossfp32 · [8192, 65536]
NVIDIA H100
84.0ms
#2 of 2
2024-10-02
Jensen-Shannon divergence lossfp32 · [8192, 131072]
NVIDIA H100
167.9ms
#2 of 2
2024-10-02

Reported · How evidence levels are derived →

Source and license

sourceavailable
revision digestsha256:a23886e9eb82a5ee86dc8da4a8b71c3d32e2af6c350b02bb1702dd8d8bdb05b1
license declaredBSD-3-Clause
license concludedBSD-3-Clause
imported2026-08-18

Best evidence level for this revision: reported

JSON