submission 554041
idanbeck · python · License unknown
Use it
Vendorable · source mirrored · license unknownView source →
No package. Vendor the mirrored source: 26 lines, June 9 Researcher Reciprocity License v1.0.
submission.py
curl "https://kernelindex.com/api/v1/implementations/kernelbot-matmul-v2-554041?include=source"interfacepython
Compatibility
measured onNVIDIA A100
declared hardwareNVIDIA A100
architecturessm_80
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 digestsha256:74547463987636c9d3e5aca64c7eef3e9c1beb60cb2d049d92e11740f7e55322
license declaredunknown
license concludedunknown
authorsidanbeck
imported2026-08-15
Techniques
Extracted from the mirrored source by pattern, never inferred. Each row cites its line.
autotune
return torch.compile(f, fullgraph=False, mode="max-autotune-no-cudagraphs")Kernel source
submission.py26 lines
#!POPCORN leaderboard matmul_v2
#!POPCORN gpu A100
import os
os.environ.setdefault("CUBLAS_WORKSPACE_CONFIG", ":4096:8")
from task import input_t, output_t
import torch
_F = None
def _build():
def f(a, b, c):
c[...] = torch.mm(a, b)
return c
try:
return torch.compile(f, fullgraph=False, mode="max-autotune-no-cudagraphs")
except Exception:
return f
@torch.inference_mode()
def custom_kernel(data: input_t) -> output_t:
global _F
torch.use_deterministic_algorithms(True)
if _F is None:
_F = _build()
a, b, c = data
return _F(a, b, 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