Megatron fused

{
  "run": {
    "kind": "BenchmarkRun",
    "spec": {
      "status": "passed",
      "timing": {
        "latencyNs": {
          "median": 2781858
        },
        "primaryStatistic": "median"
      },
      "observedAt": "2026-06-15T22:13:03.000Z",
      "measurements": [
        {
          "unit": "ns",
          "value": 2781548,
          "metric": "latency",
          "statistic": "p20"
        },
        {
          "unit": "ns",
          "value": 2783004,
          "metric": "latency",
          "statistic": "p80"
        }
      ],
      "sourceNative": {
        "source": "liger-kernel-bench",
        "metrics": {
          "ms_20": 2.781548023223877,
          "ms_50": 2.781857967376709,
          "ms_80": 2.7830042839050293
        },
        "benchmark": "megatron_cross_entropy/full",
        "externalId": "megatron_cross_entropy/full/megatron-unfused/nvidia-h100-80gb-hbm3/batch4-seq2048-vocab16384-bf16/2026-06-15-22-13-03"
      },
      "protocolDigest": "sha256:5b02ee9a2140fbc6fcc1c89800f6c177ba22718c8897cb532bf0ba91867c1e58",
      "workloadDigest": "sha256:7ca22057406780de30e12fa8d106529707acec37349c761d98dcd5aa8579da21",
      "environmentDigest": "sha256:2750ae381d6de582f1eb6b9008e0987463d19ede8af127d97d76c5ae09121393",
      "implementationDigest": "sha256:d9055243d96f775b3343dd8a494ac60fd04f4df9fdd5506871ddbce7ad70a00c"
    },
    "metadata": {
      "name": "liger-megatron-cross-entropy-full-megatron-unfused-nvidia-h100-80gb-hbm3-batch4-seq2048-vocab16384-b",
      "title": "Vocab-parallel cross-entropy, Megatron layout · megatron-unfused · full",
      "labels": {
        "liger_version": "0.8.0"
      }
    },
    "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 forward pass plus the backward 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-full",
      "title": "Liger benchmark harness · full 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"
  }
}