amd-mixture-of-experts · submission 12732
AMD DeepSeek-style MoE layer · bs = 2 · seed = 12819 · seqlen = 4096 · dexpert = 2048 · dhidden = 7168 · nroutedexperts = 32 · nsharedexperts = 1 · nexpertspertoken = 4 · fp16 · run 01a006cd-caa5…
●passedReported evidence
Not re-observed recently.
Primary measurement
6.24smean of 3
Rank 1 in its comparison group · source-native comparison · observed 2025-04-29
best ns 6233344633 · mean ns 6239639932.666667 · leaderboard score s 6.22846107348973
Identity
implementationamd-mixture-of-experts · submission 12732
projectKernelBot user 193
revisionsubmission-12732
workloadbs = 2 · seed = 12819 · seqlen = 4096 · dexpert = 2048 · dhidden = 7168 · nroutedexperts = 32 · nsharedexperts = 1 · nexpertspertoken = 4 · fp16
comparison keysha256:05d254b826f15bd6…
sourceGPU MODE KernelBot
external idsubmission/12732/benchmark/1
sha256:0fc7bf4721ec9ed213f1d6d…
Correctness
Marked passed by the source; the correctness policy was not published.
Workload
bs2
seed12819
seqlen4096
dexpert2048
dhidden7168
nroutedexperts32
nsharedexperts1
nexpertspertoken4
inputfp16 [2, 4096, 7168]
definition comparatorkernelbot_eval
Measurements
latency · max6.24 s · n=3
latency · mean6.24 s · n=3
latency · min6.23 s · n=3
latency · std5.47 ms · n=3
Protocol
harnessreference-kernels eval.py
timerwall_clock
primaryStatisticmean
correctnessReferenceproblem reference implementation
comparabilityFamilygpumode_kernelbot
Environment
gpuAMD Instinct MI300X (gfx942)
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": 3,
"latencyNs": {
"mean": 6239639933,
"maximum": 6243247098,
"minimum": 6233344633
},
"primaryStatistic": "mean"
},
"observedAt": "2025-04-29T19:23:42.925Z",
"measurements": [
{
"unit": "ns",
"value": 5471220.422775008,
"metric": "latency",
"statistic": "std",
"sampleCount": 3
}
],
"sourceNative": {
"source": "gpumode-kernelbot",
"metrics": {
"best_ns": 6233344633,
"mean_ns": 6239639932.666667,
"leaderboard_score_s": 6.22846107348973
},
"benchmark": "amd-mixture-of-experts/1",
"externalId": "submission/12732"
},
"protocolDigest": "sha256:7c66da4b5866502eb10444a7ba013d0f6bdf1620c55847945008cad2fbf98aec",
"workloadDigest": "sha256:35f375e7a77c580f0381aae4207dd30d38957ea4b0c434f5fcecdbf3ef8bd5a8",
"environmentDigest": "sha256:f722892449ed976f3e21b12df8b92b34573fc232883fec6a2f3da399a00abdd5",
"implementationDigest": "sha256:cc60d026136a3dcedcb630e2fa93c8207544ba6d8f1d3c91b9c68932a5654dfd"
},
"metadata": {
"name": "kernelbot-12732-bs2-dexpert2048-dhidden7168-nexpertspertoken4-nroutedexperts32-nsharedexperts1-seed1",
"title": "amd-mixture-of-experts · submission 12732 · case 1",
"labels": {
"torch": "2.8.0.dev20250428+rocm6.3",
"platform": "Linux-5.15.0-136-generic-x86_64-with-glibc2.35"
}
},
"apiVersion": "kernelindex.dev/v1alpha1"
},
"protocol": {
"kind": "BenchmarkProtocol",
"spec": {
"harness": {
"name": "reference-kernels eval.py",
"repository": "https://github.com/gpu-mode/reference-kernels"
},
"correctness": {
"reference": "problem reference implementation",
"comparator": "kernelbot_eval"
},
"measurement": {
"timer": "wall_clock",
"primaryStatistic": "mean"
},
"comparability": {
"notes": "Per-case wall-clock statistics as published in the KernelBot dataset. Comparable only within one leaderboard problem and GPU type; runner software versions ride each run's labels.",
"family": "gpumode_kernelbot"
}
},
"metadata": {
"name": "gpumode-kernelbot-eval",
"title": "GPU MODE KernelBot evaluation harness"
},
"apiVersion": "kernelindex.dev/v1alpha1"
},
"environment": {
"kind": "ExecutionEnvironment",
"spec": {
"hardware": {
"vendor": "amd",
"product": "AMD Instinct MI300X",
"architecture": "gfx942"
},
"software": {}
},
"metadata": {
"name": "gpumode-kernelbot-amd-instinct-mi300x",
"title": "KernelBot runner · AMD Instinct MI300X"
},
"apiVersion": "kernelindex.dev/v1alpha1"
}
}Cite this record (permalink, digest, access date)
Report an issue with this run
Published 2026-08-15 · GPU MODE KernelBot · June 9 Researcher Reciprocity License v1.0All results for AMD DeepSeek-style MoE layer →JSON