Run 01a00b1d · KernelIndex

Primary measurement

2.69msmean

Rank 1 in its comparison group · source-native comparison · observed 2026-04-30

34.52× reference

Identity

implementationflashinfer_wrapper_ad4135

projectFlashInfer-Bench baselines

revisionda915083d4c7c5e61aa3005e3d17ae488e0fc71c

workloadnum_pages = 29189 · batch_size = 64 · len_indptr = 65 · num_kv_indices = 26582 · bf16

comparison keysha256:c9b34b0ef418e546…

sourceFlashInfer-Bench

external idtrace/flashinfer_wrapper_ad4135/51393402-8fec-4b52-88ca-3cdb2270d293

sha256:b76335a801af495b57e57e7…

Correctness

comparatorflashinfer_bench_eval

max absolute error0.0009765625

max relative error769.1510009765625

resultpassed

Workload

num_pages29189

batch_size64

len_indptr65

num_kv_indices26582

qbf16 [64, 32, 128]

k_cachebf16 [29189, 64, 8, 128]

v_cachebf16 [29189, 64, 8, 128]

kv_indptrint32 [65]

kv_indicesint32 [26582]

kv_last_page_lenint32 [64]

definition comparatorsol_execbench_eval

Measurements

latency · mean2.69 ms

Protocol

harnessflashinfer-bench

timerharness_reported

primaryStatisticmean

correctnessReferencedefinition reference implementation

comparabilityFamilyflashinfer_bench_trace

Environment

gpuNVIDIA B200 (unknown)

cudaToolkit12.8

frameworkpytorch 2.9.1+cu128

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": {
          "mean": 2691617
        },
        "primaryStatistic": "mean"
      },
      "observedAt": "2026-04-30T01:31:41.029Z",
      "correctness": {
        "comparator": "sol_execbench_eval",
        "maximumAbsoluteError": 0.0009765625,
        "maximumRelativeError": 769.1510009765625
      },
      "sourceNative": {
        "source": "flashinfer-bench",
        "metrics": {
          "latency_ms": 2.6916171666666666,
          "speedup_factor": 34.52181448785777,
          "reference_latency_ms": 92.9195085
        },
        "benchmark": "gqa_paged_decode_h32_kv8_d128_ps64"
      },
      "protocolDigest": "sha256:7d44d206ad086b4bece35ce31f6531ebb11f07e0d9dcfb74c75be86d3ec02095",
      "workloadDigest": "sha256:23860ba7764f34f86697978f0270d04c60cc594cdc95d66d82dab27598d5d869",
      "environmentDigest": "sha256:af7cb68a0bb0eb9671092a620668c2600f0431301fc46a71807ea75fb7de8bce",
      "implementationDigest": "sha256:4b3e387682a1d3046578b8c6593e5e44baeea6265c8f93fe446750b194be7463"
    },
    "metadata": {
      "name": "sol-trace-gqa-paged-decode-h32-kv8-d128-ps64-51393402",
      "title": "flashinfer_wrapper_ad4135 · gqa_paged_decode_h32_kv8_d128_ps64"
    },
    "apiVersion": "kernelindex.dev/v1alpha1"
  },
  "protocol": {
    "kind": "BenchmarkProtocol",
    "spec": {
      "harness": {
        "name": "flashinfer-bench",
        "repository": "https://github.com/flashinfer-ai/flashinfer-bench"
      },
      "correctness": {
        "reference": "definition reference implementation",
        "comparator": "flashinfer_bench_eval"
      },
      "measurement": {
        "timer": "harness_reported",
        "primaryStatistic": "mean"
      },
      "comparability": {
        "family": "flashinfer_bench_trace"
      }
    },
    "metadata": {
      "name": "flashinfer-bench-trace",
      "title": "FlashInfer-Bench trace harness"
    },
    "apiVersion": "kernelindex.dev/v1alpha1"
  },
  "environment": {
    "kind": "ExecutionEnvironment",
    "spec": {
      "hardware": {
        "vendor": "nvidia",
        "product": "NVIDIA B200",
        "architecture": "unknown"
      },
      "software": {
        "framework": {
          "name": "pytorch",
          "version": "2.9.1+cu128"
        },
        "libraries": {
          "triton": "3.5.1"
        },
        "cudaToolkit": "12.8"
      }
    },
    "metadata": {
      "name": "sol-trace-nvidia-b200-2-9-1-cu128-3-5-1-12-8",
      "title": "NVIDIA B200 (trace-reported software)"
    },
    "apiVersion": "kernelindex.dev/v1alpha1"
  }
}

Cite this record (permalink, digest, access date)

Report an issue with this run

Published 2026-08-16 · FlashInfer-Bench · Apache-2.0All results for GQA paged decode h32 kv8 d128 ps64 →JSON

flashinfer / wrapperad4135