Skip to content
KernelIndex
Search⌘K

submission 31736

KernelBot user 364 · python · License unknown

Use it

Vendorable · source mirrored · license unknownView source →

No package. Vendor the mirrored source: 1 lines, June 9 Researcher Reciprocity License v1.0.

triton_v5b-2.py
curl "https://kernelindex.com/api/v1/implementations/kernelbot-amd-fp8-mm-31736?include=source"
interfacepython
Compatibility
measured onAMD Instinct MI300X
declared hardwareAMD Instinct MI300X
architecturesgfx942
dtypesfp32, fp8_e4m3fnuz

Benchmark evidence

18 measurements across 1 GPU, fastest first.

Operation / workload
Hardware
Latency
Rank
Observed
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [1024, 256]
AMD Instinct MI300X
76.5µs
#83 of 98
2025-06-09
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [1024, 512]
AMD Instinct MI300X
78.4µs
#82 of 98
2025-06-09
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [1024, 1536]
AMD Instinct MI300X
99.4µs
#75 of 98
2025-06-09
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [1024, 7168] · n=576 · seed=12346
AMD Instinct MI300X
126.9µs
#55 of 98
2025-06-09
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [1024, 7168] · n=512 · seed=6563
AMD Instinct MI300X
127.5µs
#54 of 98
2025-06-09
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [6144, 256]
AMD Instinct MI300X
139.8µs
#67 of 98
2025-06-09
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [6144, 512]
AMD Instinct MI300X
144.1µs
#68 of 98
2025-06-09
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [1024, 2048]
AMD Instinct MI300X
146.1µs
#58 of 98
2025-06-09
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [1024, 2304]
AMD Instinct MI300X
155.6µs
#57 of 98
2025-06-09
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [1024, 7168] · n=1536 · seed=8135
AMD Instinct MI300X
172.4µs
#54 of 98
2025-06-09
Show all 18 measurements ›
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [6144, 1536]
AMD Instinct MI300X
207.5µs
#54 of 98
2025-06-09
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [6144, 7168] · n=512 · seed=12341
AMD Instinct MI300X
262.8µs
#60 of 98
2025-06-09
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [6144, 7168] · n=576 · seed=9863
AMD Instinct MI300X
269.1µs
#58 of 98
2025-06-09
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [1024, 7168] · n=4608 · seed=7531
AMD Instinct MI300X
310.6µs
#64 of 98
2025-06-09
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [6144, 7168] · n=1536 · seed=6543
AMD Instinct MI300X
383.4µs
#44 of 98
2025-06-09
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [6144, 2048]
AMD Instinct MI300X
470.0µs
#42 of 98
2025-06-09
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [6144, 2304]
AMD Instinct MI300X
487.6µs
#36 of 98
2025-06-09
AMD FP8 blockwise GEMMfp32/fp8_e4m3fnuz · [6144, 7168] · n=4608 · seed=65436
AMD Instinct MI300X
809.3µs
#26 of 98
2025-06-09

Reported · How evidence levels are derived →

Source and license

sourceavailable
revision digestsha256:d99c2b481b0cb68f4af5a85e641de8ad902ff3f56ca2a3baa75aa36c6250f36f
license declaredunknown
license concludedunknown
imported2026-08-15

Kernel source

triton_v5b-2.py1 lines
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

Source code from GPU Mode and the KernelBot dataset · June 9 Researcher Reciprocity License v1.0

Best evidence level for this revision: reported

JSON