Run 01a00b1d · KernelIndex

Primary measurement

127.6µsmean

Rank 1 in its comparison group · source-native comparison · observed 2026-04-09

4.66× reference

Identity

implementationflashinfer_wrapper_96864e

projectFlashInfer-Bench baselines

revisionda915083d4c7c5e61aa3005e3d17ae488e0fc71c

workloadnum_pages = 7189 · batch_size = 3 · len_indptr = 4 · num_kv_indices = 1283 · bf16

comparison keysha256:1a4fa10468bc9b69…

sourceFlashInfer-Bench

external idtrace/flashinfer_wrapper_96864e/e7568c96-d1d0-495d-9de8-fc04df84153b

sha256:8a401acab5b81590a5bed78…

Correctness

comparatorflashinfer_bench_eval

max absolute error0.001953125

max relative error5.663188934326172

resultpassed

Workload

num_pages7189

batch_size3

len_indptr4

num_kv_indices1283

qbf16 [3, 24, 128]

k_cachebf16 [7189, 1, 8, 128]

v_cachebf16 [7189, 1, 8, 128]

kv_indptrint32 [4]

kv_indicesint32 [1283]

definition comparatorsol_execbench_eval

Measurements

latency · mean127.6 µs

Protocol

harnessflashinfer-bench

timerharness_reported

primaryStatisticmean

correctnessReferencedefinition reference implementation

comparabilityFamilyflashinfer_bench_trace

Environment

gpuNVIDIA B200 (unknown)

cudaToolkit13.0

frameworkpytorch 2.11.0+cu130

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": 127648
        },
        "primaryStatistic": "mean"
      },
      "observedAt": "2026-04-09T23:33:06.577Z",
      "correctness": {
        "comparator": "sol_execbench_eval",
        "maximumAbsoluteError": 0.001953125,
        "maximumRelativeError": 5.663188934326172
      },
      "sourceNative": {
        "source": "flashinfer-bench",
        "metrics": {
          "latency_ms": 0.1276479959487915,
          "speedup_factor": 4.663825459544317,
          "reference_latency_ms": 0.5953279733657837
        },
        "benchmark": "gqa_paged_decode_h24_kv8_d128_ps1"
      },
      "protocolDigest": "sha256:7d44d206ad086b4bece35ce31f6531ebb11f07e0d9dcfb74c75be86d3ec02095",
      "workloadDigest": "sha256:a7bde71c36ad3f9263326079b6ff7c4b988e1d37ea91e067a4914d8c991fef23",
      "environmentDigest": "sha256:83c9b6e78d493927f18878d57448d9a0d2221a70840b562455b2665a2897ba19",
      "implementationDigest": "sha256:33dbdb780e6716cd51f6c48859f9edba1b8da97fe4a128a02f1776982434e254"
    },
    "metadata": {
      "name": "sol-trace-gqa-paged-decode-h24-kv8-d128-ps1-e7568c96",
      "title": "flashinfer_wrapper_96864e · gqa_paged_decode_h24_kv8_d128_ps1"
    },
    "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": "2.11.0+cu130"
        },
        "libraries": {
          "triton": "3.6.0"
        },
        "cudaToolkit": "13.0"
      }
    },
    "metadata": {
      "name": "sol-trace-nvidia-b200-2-11-0-cu130-3-6-0-13-0",
      "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 h24 kv8 d128 ps1 →JSON

flashinfer / wrapper96864e