[
  {
    "event": "anvil_strata_prefill_timing",
    "gpu_timeline_ms": 1835,
    "host_staging_ms": 815,
    "phases": {
      "combine": {
        "ms": 1,
        "percent": 0.1
      },
      "dequant": {
        "ms": 865,
        "percent": 47.1
      },
      "embed+steps": {
        "ms": 0,
        "percent": 0.0
      },
      "gather": {
        "ms": 430,
        "percent": 23.4
      },
      "gdn": {
        "ms": 186,
        "percent": 10.1
      },
      "gdn conv+gates": {
        "ms": 1,
        "percent": 0.0
      },
      "gdn out proj": {
        "ms": 9,
        "percent": 0.5
      },
      "gdn recurrence": {
        "ms": 2,
        "percent": 0.1
      },
      "gemm down": {
        "ms": 28,
        "percent": 1.5
      },
      "gemm gate/up": {
        "ms": 28,
        "percent": 1.5
      },
      "hc read": {
        "ms": 209,
        "percent": 11.4
      },
      "host grouping": {
        "ms": 4,
        "percent": 0.2
      },
      "ple": {
        "ms": 1,
        "percent": 0.0
      },
      "qsa attn": {
        "ms": 0,
        "percent": 0.0
      },
      "qsa indexer": {
        "ms": 0,
        "percent": 0.0
      },
      "qsa proj": {
        "ms": 11,
        "percent": 0.6
      },
      "qsa select": {
        "ms": 0,
        "percent": 0.0
      },
      "router+shared": {
        "ms": 12,
        "percent": 0.7
      },
      "wait copy": {
        "ms": 48,
        "percent": 2.6
      }
    },
    "tokens": 65,
    "wall_ms": 1837
  },
  {
    "after_chunk_ms": 2,
    "chunk_setup_ms": 0,
    "chunk_wait_ms": 1,
    "event": "anvil_strata_prefill_host_timing",
    "ple_ms": 0
  },
  {
    "event": "anvil_strata_prefill_timing",
    "gpu_timeline_ms": 9358,
    "host_staging_ms": 1300,
    "phases": {
      "combine": {
        "ms": 71,
        "percent": 0.8
      },
      "dequant": {
        "ms": 4982,
        "percent": 53.2
      },
      "embed+steps": {
        "ms": 8,
        "percent": 0.1
      },
      "gather": {
        "ms": 14,
        "percent": 0.2
      },
      "gdn": {
        "ms": 109,
        "percent": 1.2
      },
      "gdn conv+gates": {
        "ms": 35,
        "percent": 0.4
      },
      "gdn out proj": {
        "ms": 74,
        "percent": 0.8
      },
      "gdn recurrence": {
        "ms": 107,
        "percent": 1.1
      },
      "gemm down": {
        "ms": 457,
        "percent": 4.9
      },
      "gemm gate/up": {
        "ms": 201,
        "percent": 2.2
      },
      "hc read": {
        "ms": 156,
        "percent": 1.7
      },
      "host grouping": {
        "ms": 2846,
        "percent": 30.4
      },
      "ple": {
        "ms": 107,
        "percent": 1.1
      },
      "qsa attn": {
        "ms": 52,
        "percent": 0.6
      },
      "qsa indexer": {
        "ms": 0,
        "percent": 0.0
      },
      "qsa proj": {
        "ms": 63,
        "percent": 0.7
      },
      "qsa select": {
        "ms": 14,
        "percent": 0.1
      },
      "router+shared": {
        "ms": 29,
        "percent": 0.3
      },
      "wait copy": {
        "ms": 33,
        "percent": 0.4
      }
    },
    "tokens": 5487,
    "wall_ms": 9364
  },
  {
    "after_chunk_ms": 5,
    "chunk_setup_ms": 5,
    "chunk_wait_ms": 13,
    "event": "anvil_strata_prefill_host_timing",
    "ple_ms": 107
  },
  {
    "event": "anvil_strata_prefill_timing",
    "gpu_timeline_ms": 2704,
    "host_staging_ms": 2732,
    "phases": {
      "combine": {
        "ms": 2,
        "percent": 0.1
      },
      "dequant": {
        "ms": 1525,
        "percent": 56.4
      },
      "embed+steps": {
        "ms": 0,
        "percent": 0.0
      },
      "gather": {
        "ms": 798,
        "percent": 29.5
      },
      "gdn": {
        "ms": 28,
        "percent": 1.0
      },
      "gdn conv+gates": {
        "ms": 2,
        "percent": 0.1
      },
      "gdn out proj": {
        "ms": 11,
        "percent": 0.4
      },
      "gdn recurrence": {
        "ms": 5,
        "percent": 0.2
      },
      "gemm down": {
        "ms": 84,
        "percent": 3.1
      },
      "gemm gate/up": {
        "ms": 68,
        "percent": 2.5
      },
      "hc read": {
        "ms": 9,
        "percent": 0.3
      },
      "host grouping": {
        "ms": 6,
        "percent": 0.2
      },
      "ple": {
        "ms": 1,
        "percent": 0.0
      },
      "qsa attn": {
        "ms": 1,
        "percent": 0.0
      },
      "qsa indexer": {
        "ms": 0,
        "percent": 0.0
      },
      "qsa proj": {
        "ms": 12,
        "percent": 0.4
      },
      "qsa select": {
        "ms": 1,
        "percent": 0.0
      },
      "router+shared": {
        "ms": 12,
        "percent": 0.4
      },
      "wait copy": {
        "ms": 140,
        "percent": 5.2
      }
    },
    "tokens": 303,
    "wall_ms": 2704
  },
  {
    "after_chunk_ms": 1,
    "chunk_setup_ms": 0,
    "chunk_wait_ms": 4,
    "event": "anvil_strata_prefill_host_timing",
    "ple_ms": 107
  },
  {
    "event": "anvil_strata_prefill_timing",
    "gpu_timeline_ms": 635,
    "host_staging_ms": 2970,
    "phases": {
      "combine": {
        "ms": 1,
        "percent": 0.2
      },
      "dequant": {
        "ms": 324,
        "percent": 51.1
      },
      "embed+steps": {
        "ms": 0,
        "percent": 0.0
      },
      "gather": {
        "ms": 113,
        "percent": 17.8
      },
      "gdn": {
        "ms": 26,
        "percent": 4.0
      },
      "gdn conv+gates": {
        "ms": 1,
        "percent": 0.1
      },
      "gdn out proj": {
        "ms": 10,
        "percent": 1.6
      },
      "gdn recurrence": {
        "ms": 3,
        "percent": 0.4
      },
      "gemm down": {
        "ms": 34,
        "percent": 5.3
      },
      "gemm gate/up": {
        "ms": 39,
        "percent": 6.1
      },
      "hc read": {
        "ms": 7,
        "percent": 1.1
      },
      "host grouping": {
        "ms": 3,
        "percent": 0.5
      },
      "ple": {
        "ms": 1,
        "percent": 0.1
      },
      "qsa attn": {
        "ms": 0,
        "percent": 0.0
      },
      "qsa indexer": {
        "ms": 0,
        "percent": 0.0
      },
      "qsa proj": {
        "ms": 11,
        "percent": 1.8
      },
      "qsa select": {
        "ms": 0,
        "percent": 0.0
      },
      "router+shared": {
        "ms": 11,
        "percent": 1.7
      },
      "wait copy": {
        "ms": 51,
        "percent": 8.0
      }
    },
    "tokens": 138,
    "wall_ms": 635
  },
  {
    "after_chunk_ms": 0,
    "chunk_setup_ms": 0,
    "chunk_wait_ms": 4,
    "event": "anvil_strata_prefill_host_timing",
    "ple_ms": 108
  },
  {
    "event": "anvil_strata_prefill_timing",
    "gpu_timeline_ms": 12522,
    "host_staging_ms": 3602,
    "phases": {
      "combine": {
        "ms": 97,
        "percent": 0.8
      },
      "dequant": {
        "ms": 7322,
        "percent": 58.5
      },
      "embed+steps": {
        "ms": 19,
        "percent": 0.2
      },
      "gather": {
        "ms": 16,
        "percent": 0.1
      },
      "gdn": {
        "ms": 147,
        "percent": 1.2
      },
      "gdn conv+gates": {
        "ms": 58,
        "percent": 0.5
      },
      "gdn out proj": {
        "ms": 112,
        "percent": 0.9
      },
      "gdn recurrence": {
        "ms": 156,
        "percent": 1.2
      },
      "gemm down": {
        "ms": 554,
        "percent": 4.4
      },
      "gemm gate/up": {
        "ms": 332,
        "percent": 2.7
      },
      "hc read": {
        "ms": 236,
        "percent": 1.9
      },
      "host grouping": {
        "ms": 3236,
        "percent": 25.8
      },
      "ple": {
        "ms": 5,
        "percent": 0.0
      },
      "qsa attn": {
        "ms": 74,
        "percent": 0.6
      },
      "qsa indexer": {
        "ms": 0,
        "percent": 0.0
      },
      "qsa proj": {
        "ms": 82,
        "percent": 0.7
      },
      "qsa select": {
        "ms": 19,
        "percent": 0.2
      },
      "router+shared": {
        "ms": 33,
        "percent": 0.3
      },
      "wait copy": {
        "ms": 22,
        "percent": 0.2
      }
    },
    "tokens": 7410,
    "wall_ms": 12532
  },
  {
    "after_chunk_ms": 9,
    "chunk_setup_ms": 5,
    "chunk_wait_ms": 13,
    "event": "anvil_strata_prefill_host_timing",
    "ple_ms": 108
  },
  {
    "event": "anvil_strata_prefill_timing",
    "gpu_timeline_ms": 10277,
    "host_staging_ms": 4061,
    "phases": {
      "combine": {
        "ms": 91,
        "percent": 0.9
      },
      "dequant": {
        "ms": 5764,
        "percent": 56.1
      },
      "embed+steps": {
        "ms": 5,
        "percent": 0.0
      },
      "gather": {
        "ms": 15,
        "percent": 0.1
      },
      "gdn": {
        "ms": 124,
        "percent": 1.2
      },
      "gdn conv+gates": {
        "ms": 34,
        "percent": 0.3
      },
      "gdn out proj": {
        "ms": 77,
        "percent": 0.8
      },
      "gdn recurrence": {
        "ms": 132,
        "percent": 1.3
      },
      "gemm down": {
        "ms": 609,
        "percent": 5.9
      },
      "gemm gate/up": {
        "ms": 181,
        "percent": 1.8
      },
      "hc read": {
        "ms": 153,
        "percent": 1.5
      },
      "host grouping": {
        "ms": 2851,
        "percent": 27.7
      },
      "ple": {
        "ms": 5,
        "percent": 0.1
      },
      "qsa attn": {
        "ms": 74,
        "percent": 0.7
      },
      "qsa indexer": {
        "ms": 0,
        "percent": 0.0
      },
      "qsa proj": {
        "ms": 78,
        "percent": 0.8
      },
      "qsa select": {
        "ms": 19,
        "percent": 0.2
      },
      "router+shared": {
        "ms": 33,
        "percent": 0.3
      },
      "wait copy": {
        "ms": 30,
        "percent": 0.3
      }
    },
    "tokens": 7412,
    "wall_ms": 10285
  },
  {
    "after_chunk_ms": 7,
    "chunk_setup_ms": 5,
    "chunk_wait_ms": 15,
    "event": "anvil_strata_prefill_host_timing",
    "ple_ms": 108
  },
  {
    "event": "anvil_strata_prefill_timing",
    "gpu_timeline_ms": 10573,
    "host_staging_ms": 4446,
    "phases": {
      "combine": {
        "ms": 95,
        "percent": 0.9
      },
      "dequant": {
        "ms": 6711,
        "percent": 63.5
      },
      "embed+steps": {
        "ms": 30,
        "percent": 0.3
      },
      "gather": {
        "ms": 16,
        "percent": 0.2
      },
      "gdn": {
        "ms": 124,
        "percent": 1.2
      },
      "gdn conv+gates": {
        "ms": 37,
        "percent": 0.3
      },
      "gdn out proj": {
        "ms": 79,
        "percent": 0.7
      },
      "gdn recurrence": {
        "ms": 133,
        "percent": 1.3
      },
      "gemm down": {
        "ms": 324,
        "percent": 3.1
      },
      "gemm gate/up": {
        "ms": 180,
        "percent": 1.7
      },
      "hc read": {
        "ms": 192,
        "percent": 1.8
      },
      "host grouping": {
        "ms": 2412,
        "percent": 22.8
      },
      "ple": {
        "ms": 5,
        "percent": 0.1
      },
      "qsa attn": {
        "ms": 74,
        "percent": 0.7
      },
      "qsa indexer": {
        "ms": 0,
        "percent": 0.0
      },
      "qsa proj": {
        "ms": 80,
        "percent": 0.8
      },
      "qsa select": {
        "ms": 19,
        "percent": 0.2
      },
      "router+shared": {
        "ms": 33,
        "percent": 0.3
      },
      "wait copy": {
        "ms": 28,
        "percent": 0.3
      }
    },
    "tokens": 7413,
    "wall_ms": 10581
  },
  {
    "after_chunk_ms": 7,
    "chunk_setup_ms": 30,
    "chunk_wait_ms": 15,
    "event": "anvil_strata_prefill_host_timing",
    "ple_ms": 108
  }
]
