torch.compile (inductor)
RegNet · batch_size = 8 · image_width = 224 · image_height = 224 · input_channels = 3 · fp32 · run 01a03f0e-f854…
●passedReported evidence
Not re-observed recently.
Primary measurement
1.60ms±0.00 · mean of 100
Rank 1 in its comparison group · source-native comparison · observed 2026-03-05
max ms 1.66 · min ms 1.59 · std ms 0.0144 · mean ms 1.6
Identity
implementationtorch.compile (inductor)
projectPyTorch
revisionunknown
workloadbatch_size = 8 · image_width = 224 · image_height = 224 · input_channels = 3 · fp32
comparison keysha256:6bda1acf8de08c16…
sourceKernelBench baseline timings
external idH100_PCIe_LambdaLabs/torch-compile-inductor/level3/27_RegNet.py
sha256:ec60d317b60d866d845552c…
Correctness
Marked passed by the source; the correctness policy was not published.
Workload
batch_size8
image_width224
image_height224
input_channels3
input_1fp32 [8, 3, 224, 224]
definition comparatornot_asserted
Measurements
latency · max1.66 ms · n=100
latency · mean1.60 ms · n=100
latency · min1.59 ms · n=100
latency · std14.4 µs
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": 1600000,
"maximum": 1660000,
"minimum": 1590000,
"confidence95": [
1597178,
1602822
]
},
"primaryStatistic": "mean"
},
"observedAt": "2026-03-05T08:38:15.000Z",
"measurements": [
{
"unit": "ns",
"value": 14400,
"metric": "latency",
"statistic": "std"
}
],
"sourceNative": {
"source": "kernelbench",
"metrics": {
"max_ms": 1.66,
"min_ms": 1.59,
"std_ms": 0.0144,
"mean_ms": 1.6
},
"benchmark": "H100_PCIe_LambdaLabs/baseline_time_torch_compile_inductor_default.json",
"externalId": "H100_PCIe_LambdaLabs/torch-compile-inductor/level3/27_RegNet.py"
},
"protocolDigest": "sha256:547725b033a9911eae0c5ac362344a85aed824894ed4ab476af945b9bb491ea6",
"workloadDigest": "sha256:988fe9404aafac8f665168fcd6b58352d59729296f21bfcd8a5c6d2c07bbe14c",
"environmentDigest": "sha256:89f9b28b06758f22a1df331fea58bfde2c9cc0f3627d1829509743ea538f0011",
"implementationDigest": "sha256:8bc8273ac778cf7c900fb08a31ed0aa1b64d125af4f24dea85a8e22927e8574d"
},
"metadata": {
"name": "kernelbench-h100-pcie-lambdalabs-torch-compile-inductor-l3-27-regnet",
"title": "RegNet · torch.compile (inductor) · Lambda Labs"
},
"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": "PCIe",
"architecture": "sm_90"
},
"software": {}
},
"metadata": {
"name": "kernelbench-h100-pcie-lambdalabs",
"title": "KernelBench host · H100 PCIe (Lambda Labs)"
},
"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 RegNet →JSON