torch.compile (inductor)
4D tensor matrix multiplication · b = 8 · i = 256 · j = 512 · k = 768 · l = 256 · fp32 · run 01a03f0e-f6de…
●passedReported evidence
Not re-observed recently.
Primary measurement
10.9ms±0.21 · mean of 100
Rank 1 in its comparison group · source-native comparison · observed 2026-03-05
max ms 15.6 · min ms 10 · std ms 1.07 · mean ms 10.9
Identity
implementationtorch.compile (inductor)
projectPyTorch
revisionunknown
workloadb = 8 · i = 256 · j = 512 · k = 768 · l = 256 · fp32
comparison keysha256:ad987457102508e0…
sourceKernelBench baseline timings
external idH100_Modal/torch-compile-inductor/level1/11_4D_tensor_matrix_multiplication.py
sha256:05015e7427e70b38896285e…
Correctness
Marked passed by the source; the correctness policy was not published.
Workload
b8
i256
j512
k768
l256
afp32 [8, 256, 512, 256]
bfp32 [256, 768]
definition comparatornot_asserted
Measurements
latency · max15.6 ms · n=100
latency · mean10.9 ms · n=100
latency · min10.0 ms · n=100
latency · std1.07 ms
Protocol
harnessKernelBench timing scripts
timercuda_events
primaryStatisticmean
comparabilityFamilykernelbench_baseline_timing
Environment
gpuNVIDIA H100 (sm_90)
Artifacts
No artifacts published with this run.
Replications and notes
No attestations yet.
Community attestations. They never change the evidence level; only a KernelIndex-controlled rerun does.
Add a reproduction or note
Canonical manifest
Show manifestHide manifest
{
"run": {
"kind": "BenchmarkRun",
"spec": {
"status": "passed",
"timing": {
"samples": 100,
"latencyNs": {
"mean": 10900000,
"maximum": 15600000,
"minimum": 10000000,
"confidence95": [
10690280,
11109720
]
},
"primaryStatistic": "mean"
},
"observedAt": "2026-03-05T08:38:15.000Z",
"measurements": [
{
"unit": "ns",
"value": 1070000,
"metric": "latency",
"statistic": "std"
}
],
"sourceNative": {
"source": "kernelbench",
"metrics": {
"max_ms": 15.6,
"min_ms": 10,
"std_ms": 1.07,
"mean_ms": 10.9
},
"benchmark": "H100_Modal/baseline_time_torch_compile_inductor_default.json",
"externalId": "H100_Modal/torch-compile-inductor/level1/11_4D_tensor_matrix_multiplication.py"
},
"protocolDigest": "sha256:547725b033a9911eae0c5ac362344a85aed824894ed4ab476af945b9bb491ea6",
"workloadDigest": "sha256:6df7702acbf46673dbde46d544197b135025e0bed166539ba20f7f216427c98a",
"environmentDigest": "sha256:2750ae381d6de582f1eb6b9008e0987463d19ede8af127d97d76c5ae09121393",
"implementationDigest": "sha256:dc572fba73a1c59a093f530b47b6ada6530fb891d24d38fc9ee1a41db65f729e"
},
"metadata": {
"name": "kernelbench-h100-modal-torch-compile-inductor-l1-11-4d-tensor-matrix-multiplication",
"title": "4D tensor matrix multiplication · torch.compile (inductor) · Modal"
},
"apiVersion": "kernelindex.dev/v1alpha1"
},
"protocol": {
"kind": "BenchmarkProtocol",
"spec": {
"harness": {
"name": "KernelBench timing scripts",
"repository": "https://github.com/ScalingIntelligence/KernelBench"
},
"measurement": {
"timer": "cuda_events",
"primaryStatistic": "mean"
},
"comparability": {
"notes": "Mean of 100 timed forward passes (CUDA events, warm-up excluded) of the reference module on fixed inputs; the 95% interval is the mean's, from the reported standard deviation. The upstream JSON records no torch or CUDA version. Comparable only within one problem and one timing host.",
"family": "kernelbench_baseline_timing"
}
},
"metadata": {
"name": "kernelbench-timing-v1",
"title": "KernelBench baseline timing"
},
"apiVersion": "kernelindex.dev/v1alpha1"
},
"environment": {
"kind": "ExecutionEnvironment",
"spec": {
"hardware": {
"vendor": "nvidia",
"product": "NVIDIA H100",
"formFactor": "80GB HBM3",
"architecture": "sm_90"
},
"software": {}
},
"metadata": {
"name": "kernelbench-h100-modal",
"title": "KernelBench host · H100 80GB HBM3 (Modal)"
},
"apiVersion": "kernelindex.dev/v1alpha1"
}
}Cite this record (permalink, digest, access date)
Report an issue with this run
Published 2026-08-26 · KernelBench baseline timings · MITAll results for 4D tensor matrix multiplication →JSON