submission 458326
Nick Nuon · python · License unknown
Use it
Vendorable · source mirrored · license unknownView source →
No package. Vendor the mirrored source: 12 lines, June 9 Researcher Reciprocity License v1.0.
matmul.py
curl "https://kernelindex.com/api/v1/implementations/kernelbot-matmul-v2-458326?include=source"interfacepython
Compatibility
measured onNVIDIA B200
declared hardwareNVIDIA B200
architecturessm_100
dtypesfp16
Benchmark evidence
1 measurement across 1 GPU, fastest first.
Operation / workload
Hardware
Latency
Rank
Observed
Reported · How evidence levels are derived →
Source and license
sourceavailable
revision digestsha256:1526de7b3de033fd3a7c9069ccdf1dd072a66d63b5c98c724499dee47579936b
license declaredunknown
license concludedunknown
authorsNick Nuon
imported2026-08-15
Kernel source
matmul.py12 lines
import os
# Must be set before the first cuBLAS GEMM for determinism under torch deterministic mode
os.environ.setdefault("CUBLAS_WORKSPACE_CONFIG", ":4096:8")
import torch
@torch.inference_mode()
def custom_kernel(data):
a, b, c = data[0], data[1], data[2]
c.copy_(a @ b)
return c
Source code from GPU Mode and the KernelBot dataset · June 9 Researcher Reciprocity License v1.0
Best evidence level for this revision: reported
JSON