Run 01a0202b · KernelIndex

Primary measurement

84.6msmedian

Rank 2 in its comparison group · source-native comparison · observed 2025-05-27

Compare with #1 →

ms 20 84.55020904541016 · ms 50 84.55020904541016 · ms 80 84.55020904541016

Identity

implementationPyTorch

projectPyTorch

revisionunknown

workloadseq = 1024 · batch = 2 · heads = 8 · hidden = 512 · dilation = 2 · kernel_size = 7 · fp32

comparison keysha256:b614dce1fec1aaee…

sourceLiger-Kernel benchmarks

external idfused_neighborhood_attention/forward/torch/nvidia-h100-80gb-hbm3/batch2-dilation2-heads8-hidden512-kernel_size7-seq1024-fp32/2025-05-27-19-26-55

sha256:cbe32a64d86325890e554e7…

Correctness

Marked passed by the source; the correctness policy was not published.

Workload

seq1024

batch2

heads8

hidden512

dilation2

kernel_size7

k_biasfp32 [512]

q_biasfp32 [512]

v_biasfp32 [512]

k_weightfp32 [512, 512]

out_biasfp32 [512]

q_weightfp32 [512, 512]

v_weightfp32 [512, 512]

out_weightfp32 [512, 512]

hidden_statesfp32 [2, 1024, 512]

definition comparatornot_asserted

Measurements

latency · median84.6 ms

latency · p2084.6 ms

latency · p8084.6 ms

Protocol

harnessLiger-Kernel benchmark scripts

timercuda_events

primaryStatisticmedian

comparabilityFamilyliger_kernel_bench

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 manifest

{
  "run": {
    "kind": "BenchmarkRun",
    "spec": {
      "status": "passed",
      "timing": {
        "latencyNs": {
          "median": 84550209
        },
        "primaryStatistic": "median"
      },
      "observedAt": "2025-05-27T19:26:55.000Z",
      "measurements": [
        {
          "unit": "ns",
          "value": 84550209,
          "metric": "latency",
          "statistic": "p20"
        },
        {
          "unit": "ns",
          "value": 84550209,
          "metric": "latency",
          "statistic": "p80"
        }
      ],
      "sourceNative": {
        "source": "liger-kernel-bench",
        "metrics": {
          "ms_20": 84.55020904541016,
          "ms_50": 84.55020904541016,
          "ms_80": 84.55020904541016
        },
        "benchmark": "fused_neighborhood_attention/forward",
        "externalId": "fused_neighborhood_attention/forward/torch/nvidia-h100-80gb-hbm3/batch2-dilation2-heads8-hidden512-kernel_size7-seq1024-fp32/2025-05-27-19-26-55"
      },
      "protocolDigest": "sha256:a79eaf4df66c97c6adc1fb2d09df78682155d1ea86652ae0b70199875bde7747",
      "workloadDigest": "sha256:13f03284589f40d86af76683c14ff76e18c4a9335cabceb439a4e350b4f89fb0",
      "environmentDigest": "sha256:2750ae381d6de582f1eb6b9008e0987463d19ede8af127d97d76c5ae09121393",
      "implementationDigest": "sha256:0c3e4e2a4849358a1d5183daed7f408a3422f6cc411cc6bd7d4ebeead8e57e00"
    },
    "metadata": {
      "name": "liger-fused-neighborhood-attention-forward-torch-nvidia-h100-80gb-hbm3-batch2-dilation2-heads8-hidde",
      "title": "Fused neighborhood attention · torch · forward",
      "labels": {
        "liger_version": "0.5.10"
      }
    },
    "apiVersion": "kernelindex.dev/v1alpha1"
  },
  "protocol": {
    "kind": "BenchmarkProtocol",
    "spec": {
      "harness": {
        "name": "Liger-Kernel benchmark scripts",
        "repository": "https://github.com/linkedin/Liger-Kernel"
      },
      "measurement": {
        "timer": "cuda_events",
        "primaryStatistic": "median"
      },
      "comparability": {
        "notes": "Timed: the module forward pass, via triton.testing.do_bench with quantiles 0.5/0.2/0.8 (median with p20/p80 spread). The upstream CSV records no CUDA, driver, or torch version; the Liger release rides each run's labels. Comparable only within one kernel, workload, timed pass, and GPU.",
        "family": "liger_kernel_bench"
      }
    },
    "metadata": {
      "name": "liger-bench-do-bench-forward",
      "title": "Liger benchmark harness · forward pass"
    },
    "apiVersion": "kernelindex.dev/v1alpha1"
  },
  "environment": {
    "kind": "ExecutionEnvironment",
    "spec": {
      "hardware": {
        "vendor": "nvidia",
        "product": "NVIDIA H100",
        "formFactor": "80GB HBM3",
        "architecture": "sm_90"
      },
      "software": {}
    },
    "metadata": {
      "name": "liger-bench-nvidia-h100-80gb-hbm3",
      "title": "Liger benchmark host · H100 80GB HBM3"
    },
    "apiVersion": "kernelindex.dev/v1alpha1"
  }
}

Cite this record (permalink, digest, access date)

Report an issue with this run

Published 2026-08-20 · Liger-Kernel benchmarks · BSD-2-ClauseAll results for Fused neighborhood attention →JSON