Run 01a02032 · KernelIndex

Primary measurement

4.94msmean

Rank 4 in its comparison group · source-native comparison · observed 2026-03-28

Compare with #1 →

6.89× reference

Identity

implementationgpt-o3_triton_b12b97

projectgpt-o3

revisionda915083d4c7c5e61aa3005e3d17ae488e0fc71c

workloadnum_pages = 35359 · batch_size = 64 · len_indptr = 65 · num_kv_indices = 35343 · bf16

comparison keysha256:0046a731f4a24b3c…

sourceFlashInfer-Bench

sha256:85c87d0c244d8328b23b614…

Correctness

comparatorflashinfer_bench_eval

max absolute error0.0009765625

max relative error0.025537382811307907

resultpassed

Workload

num_pages35359

batch_size64

len_indptr65

num_kv_indices35343

qbf16 [64, 32, 128]

k_cachebf16 [35359, 1, 4, 128]

v_cachebf16 [35359, 1, 4, 128]

kv_indptrint32 [65]

kv_indicesint32 [35343]

definition comparatorsol_execbench_eval

Measurements

latency · mean4.94 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

logssha256:78e7078cb6b34029… · text/plain · unavailable

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": 4936187
        },
        "primaryStatistic": "mean"
      },
      "evidence": {
        "logs": {
          "uri": "kernelindex:artifact/sha256:78e7078cb6b34029b3c54573e2621a41c5953e9f5f65b6f20a4330ab56eda92e",
          "digest": "sha256:78e7078cb6b34029b3c54573e2621a41c5953e9f5f65b6f20a4330ab56eda92e"
        }
      },
      "observedAt": "2026-03-28T14:46:53.182Z",
      "correctness": {
        "comparator": "sol_execbench_eval",
        "maximumAbsoluteError": 0.0009765625,
        "maximumRelativeError": 0.025537382811307907
      },
      "sourceNative": {
        "source": "flashinfer-bench",
        "metrics": {
          "latency_ms": 4.936186710993449,
          "speedup_factor": 6.888909693733068,
          "reference_latency_ms": 34.004944483439125
        },
        "benchmark": "gqa_paged_decode_h32_kv4_d128_ps1"
      },
      "protocolDigest": "sha256:7d44d206ad086b4bece35ce31f6531ebb11f07e0d9dcfb74c75be86d3ec02095",
      "workloadDigest": "sha256:6c034ef7a44c27e01a02c3047246f7522c88b3e3e104511abf14f015fe03e20d",
      "environmentDigest": "sha256:af7cb68a0bb0eb9671092a620668c2600f0431301fc46a71807ea75fb7de8bce",
      "implementationDigest": "sha256:806a80269911f1ea48e10e81d53090bb50b42e8db650c6eebac66fc60534ed90"
    },
    "metadata": {
      "name": "sol-trace-gqa-paged-decode-h32-kv4-d128-ps1-eaae4f8a",
      "title": "gpt-o3_triton_b12b97 · gqa_paged_decode_h32_kv4_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": "2.9.1+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 2026-08-20 · FlashInfer-Bench · Apache-2.0All results for GQA paged decode h32 kv4 d128 ps1 →JSON

gpt-o3 / tritonb12b97