{
  "tile_sizes": {
    "M": 32,
    "K": 64,
    "N": 32
  },
  "engines": [
    "pe_dma",
    "pe_fetch_store",
    "pe_gemm",
    "pe_math"
  ],
  "stages": [
    "DMA_READ",
    "DMA_WRITE",
    "FETCH",
    "STORE",
    "GEMM",
    "MATH"
  ],
  "variants": [
    "ref_ref",
    "load_ref",
    "load_load"
  ],
  "rows": [
    {
      "M": 32,
      "K": 32,
      "N": 32,
      "variant": "ref_ref",
      "flops": 65536,
      "bytes_hbm": 6144,
      "arith_intensity": 10.666666666666666,
      "tile_count_expected": 1,
      "sim_wall_clock_s": 1.64,
      "engines": {
        "pe_dma": {
          "occupancy_ns": 16594.192,
          "wall_ns": 15271.192000000001,
          "record_count": 576
        },
        "pe_fetch_store": {
          "occupancy_ns": 8704.0,
          "wall_ns": 7624.0,
          "record_count": 320
        },
        "pe_gemm": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "pe_math": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "stages": {
        "DMA_READ": {
          "occupancy_ns": 15250.192000000001,
          "wall_ns": 15250.192000000001,
          "record_count": 512
        },
        "DMA_WRITE": {
          "occupancy_ns": 1344.0,
          "wall_ns": 1344.0,
          "record_count": 64
        },
        "FETCH": {
          "occupancy_ns": 8192.0,
          "wall_ns": 7616.0,
          "record_count": 256
        },
        "STORE": {
          "occupancy_ns": 512.0,
          "wall_ns": 512.0,
          "record_count": 64
        },
        "GEMM": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "MATH": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "pe_window_ns": 15327.576000000001,
      "composite_window_ns": 15327.576000000001
    },
    {
      "M": 32,
      "K": 32,
      "N": 32,
      "variant": "load_ref",
      "flops": 65536,
      "bytes_hbm": 6144,
      "arith_intensity": 10.666666666666666,
      "tile_count_expected": 1,
      "sim_wall_clock_s": 0.696,
      "engines": {
        "pe_dma": {
          "occupancy_ns": 16594.192,
          "wall_ns": 15271.192000000001,
          "record_count": 576
        },
        "pe_fetch_store": {
          "occupancy_ns": 8704.0,
          "wall_ns": 7624.0,
          "record_count": 320
        },
        "pe_gemm": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "pe_math": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "stages": {
        "DMA_READ": {
          "occupancy_ns": 15250.192000000001,
          "wall_ns": 15250.192000000001,
          "record_count": 512
        },
        "DMA_WRITE": {
          "occupancy_ns": 1344.0,
          "wall_ns": 1344.0,
          "record_count": 64
        },
        "FETCH": {
          "occupancy_ns": 8192.0,
          "wall_ns": 7616.0,
          "record_count": 256
        },
        "STORE": {
          "occupancy_ns": 512.0,
          "wall_ns": 512.0,
          "record_count": 64
        },
        "GEMM": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "MATH": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "pe_window_ns": 15327.576000000001,
      "composite_window_ns": 15327.576000000001
    },
    {
      "M": 32,
      "K": 32,
      "N": 32,
      "variant": "load_load",
      "flops": 65536,
      "bytes_hbm": 6144,
      "arith_intensity": 10.666666666666666,
      "tile_count_expected": 1,
      "sim_wall_clock_s": 0.605,
      "engines": {
        "pe_dma": {
          "occupancy_ns": 16594.192,
          "wall_ns": 15271.192000000001,
          "record_count": 576
        },
        "pe_fetch_store": {
          "occupancy_ns": 8704.0,
          "wall_ns": 7624.0,
          "record_count": 320
        },
        "pe_gemm": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "pe_math": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "stages": {
        "DMA_READ": {
          "occupancy_ns": 15250.192000000001,
          "wall_ns": 15250.192000000001,
          "record_count": 512
        },
        "DMA_WRITE": {
          "occupancy_ns": 1344.0,
          "wall_ns": 1344.0,
          "record_count": 64
        },
        "FETCH": {
          "occupancy_ns": 8192.0,
          "wall_ns": 7616.0,
          "record_count": 256
        },
        "STORE": {
          "occupancy_ns": 512.0,
          "wall_ns": 512.0,
          "record_count": 64
        },
        "GEMM": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "MATH": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "pe_window_ns": 15327.576000000001,
      "composite_window_ns": 15327.576000000001
    },
    {
      "M": 32,
      "K": 64,
      "N": 32,
      "variant": "ref_ref",
      "flops": 131072,
      "bytes_hbm": 10240,
      "arith_intensity": 12.8,
      "tile_count_expected": 1,
      "sim_wall_clock_s": 0.599,
      "engines": {
        "pe_dma": {
          "occupancy_ns": 16594.192,
          "wall_ns": 15271.192000000001,
          "record_count": 576
        },
        "pe_fetch_store": {
          "occupancy_ns": 8704.0,
          "wall_ns": 7624.0,
          "record_count": 320
        },
        "pe_gemm": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "pe_math": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "stages": {
        "DMA_READ": {
          "occupancy_ns": 15250.192000000001,
          "wall_ns": 15250.192000000001,
          "record_count": 512
        },
        "DMA_WRITE": {
          "occupancy_ns": 1344.0,
          "wall_ns": 1344.0,
          "record_count": 64
        },
        "FETCH": {
          "occupancy_ns": 8192.0,
          "wall_ns": 7616.0,
          "record_count": 256
        },
        "STORE": {
          "occupancy_ns": 512.0,
          "wall_ns": 512.0,
          "record_count": 64
        },
        "GEMM": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "MATH": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "pe_window_ns": 15327.576000000001,
      "composite_window_ns": 15327.576000000001
    },
    {
      "M": 32,
      "K": 64,
      "N": 32,
      "variant": "load_ref",
      "flops": 131072,
      "bytes_hbm": 10240,
      "arith_intensity": 12.8,
      "tile_count_expected": 1,
      "sim_wall_clock_s": 0.622,
      "engines": {
        "pe_dma": {
          "occupancy_ns": 16594.192,
          "wall_ns": 15271.192000000001,
          "record_count": 576
        },
        "pe_fetch_store": {
          "occupancy_ns": 8704.0,
          "wall_ns": 7624.0,
          "record_count": 320
        },
        "pe_gemm": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "pe_math": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "stages": {
        "DMA_READ": {
          "occupancy_ns": 15250.192000000001,
          "wall_ns": 15250.192000000001,
          "record_count": 512
        },
        "DMA_WRITE": {
          "occupancy_ns": 1344.0,
          "wall_ns": 1344.0,
          "record_count": 64
        },
        "FETCH": {
          "occupancy_ns": 8192.0,
          "wall_ns": 7616.0,
          "record_count": 256
        },
        "STORE": {
          "occupancy_ns": 512.0,
          "wall_ns": 512.0,
          "record_count": 64
        },
        "GEMM": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "MATH": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "pe_window_ns": 15327.576000000001,
      "composite_window_ns": 15327.576000000001
    },
    {
      "M": 32,
      "K": 64,
      "N": 32,
      "variant": "load_load",
      "flops": 131072,
      "bytes_hbm": 10240,
      "arith_intensity": 12.8,
      "tile_count_expected": 1,
      "sim_wall_clock_s": 0.63,
      "engines": {
        "pe_dma": {
          "occupancy_ns": 16594.192,
          "wall_ns": 15271.192000000001,
          "record_count": 576
        },
        "pe_fetch_store": {
          "occupancy_ns": 8704.0,
          "wall_ns": 7624.0,
          "record_count": 320
        },
        "pe_gemm": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "pe_math": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "stages": {
        "DMA_READ": {
          "occupancy_ns": 15250.192000000001,
          "wall_ns": 15250.192000000001,
          "record_count": 512
        },
        "DMA_WRITE": {
          "occupancy_ns": 1344.0,
          "wall_ns": 1344.0,
          "record_count": 64
        },
        "FETCH": {
          "occupancy_ns": 8192.0,
          "wall_ns": 7616.0,
          "record_count": 256
        },
        "STORE": {
          "occupancy_ns": 512.0,
          "wall_ns": 512.0,
          "record_count": 64
        },
        "GEMM": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "MATH": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "pe_window_ns": 15327.576000000001,
      "composite_window_ns": 15327.576000000001
    },
    {
      "M": 32,
      "K": 128,
      "N": 32,
      "variant": "ref_ref",
      "flops": 262144,
      "bytes_hbm": 18432,
      "arith_intensity": 14.222222222222221,
      "tile_count_expected": 2,
      "sim_wall_clock_s": 0.618,
      "engines": {
        "pe_dma": {
          "occupancy_ns": 16594.192,
          "wall_ns": 15271.192000000001,
          "record_count": 576
        },
        "pe_fetch_store": {
          "occupancy_ns": 8704.0,
          "wall_ns": 7624.0,
          "record_count": 320
        },
        "pe_gemm": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "pe_math": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "stages": {
        "DMA_READ": {
          "occupancy_ns": 15250.192000000001,
          "wall_ns": 15250.192000000001,
          "record_count": 512
        },
        "DMA_WRITE": {
          "occupancy_ns": 1344.0,
          "wall_ns": 1344.0,
          "record_count": 64
        },
        "FETCH": {
          "occupancy_ns": 8192.0,
          "wall_ns": 7616.0,
          "record_count": 256
        },
        "STORE": {
          "occupancy_ns": 512.0,
          "wall_ns": 512.0,
          "record_count": 64
        },
        "GEMM": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "MATH": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "pe_window_ns": 15327.576000000001,
      "composite_window_ns": 15327.576000000001
    },
    {
      "M": 32,
      "K": 128,
      "N": 32,
      "variant": "load_ref",
      "flops": 262144,
      "bytes_hbm": 18432,
      "arith_intensity": 14.222222222222221,
      "tile_count_expected": 2,
      "sim_wall_clock_s": 0.642,
      "engines": {
        "pe_dma": {
          "occupancy_ns": 16594.192,
          "wall_ns": 15271.192000000001,
          "record_count": 576
        },
        "pe_fetch_store": {
          "occupancy_ns": 8704.0,
          "wall_ns": 7624.0,
          "record_count": 320
        },
        "pe_gemm": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "pe_math": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "stages": {
        "DMA_READ": {
          "occupancy_ns": 15250.192000000001,
          "wall_ns": 15250.192000000001,
          "record_count": 512
        },
        "DMA_WRITE": {
          "occupancy_ns": 1344.0,
          "wall_ns": 1344.0,
          "record_count": 64
        },
        "FETCH": {
          "occupancy_ns": 8192.0,
          "wall_ns": 7616.0,
          "record_count": 256
        },
        "STORE": {
          "occupancy_ns": 512.0,
          "wall_ns": 512.0,
          "record_count": 64
        },
        "GEMM": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "MATH": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "pe_window_ns": 15327.576000000001,
      "composite_window_ns": 15327.576000000001
    },
    {
      "M": 32,
      "K": 128,
      "N": 32,
      "variant": "load_load",
      "flops": 262144,
      "bytes_hbm": 18432,
      "arith_intensity": 14.222222222222221,
      "tile_count_expected": 2,
      "sim_wall_clock_s": 0.769,
      "engines": {
        "pe_dma": {
          "occupancy_ns": 16594.192,
          "wall_ns": 15271.192000000001,
          "record_count": 576
        },
        "pe_fetch_store": {
          "occupancy_ns": 8704.0,
          "wall_ns": 7624.0,
          "record_count": 320
        },
        "pe_gemm": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "pe_math": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "stages": {
        "DMA_READ": {
          "occupancy_ns": 15250.192000000001,
          "wall_ns": 15250.192000000001,
          "record_count": 512
        },
        "DMA_WRITE": {
          "occupancy_ns": 1344.0,
          "wall_ns": 1344.0,
          "record_count": 64
        },
        "FETCH": {
          "occupancy_ns": 8192.0,
          "wall_ns": 7616.0,
          "record_count": 256
        },
        "STORE": {
          "occupancy_ns": 512.0,
          "wall_ns": 512.0,
          "record_count": 64
        },
        "GEMM": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "MATH": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "pe_window_ns": 15327.576000000001,
      "composite_window_ns": 15327.576000000001
    },
    {
      "M": 32,
      "K": 128,
      "N": 128,
      "variant": "ref_ref",
      "flops": 1048576,
      "bytes_hbm": 49152,
      "arith_intensity": 21.333333333333332,
      "tile_count_expected": 8,
      "sim_wall_clock_s": 0.649,
      "engines": {
        "pe_dma": {
          "occupancy_ns": 16594.192,
          "wall_ns": 15271.192000000001,
          "record_count": 576
        },
        "pe_fetch_store": {
          "occupancy_ns": 8704.0,
          "wall_ns": 7624.0,
          "record_count": 320
        },
        "pe_gemm": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "pe_math": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "stages": {
        "DMA_READ": {
          "occupancy_ns": 15250.192000000001,
          "wall_ns": 15250.192000000001,
          "record_count": 512
        },
        "DMA_WRITE": {
          "occupancy_ns": 1344.0,
          "wall_ns": 1344.0,
          "record_count": 64
        },
        "FETCH": {
          "occupancy_ns": 8192.0,
          "wall_ns": 7616.0,
          "record_count": 256
        },
        "STORE": {
          "occupancy_ns": 512.0,
          "wall_ns": 512.0,
          "record_count": 64
        },
        "GEMM": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "MATH": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "pe_window_ns": 15327.576000000001,
      "composite_window_ns": 15327.576000000001
    },
    {
      "M": 32,
      "K": 128,
      "N": 128,
      "variant": "load_ref",
      "flops": 1048576,
      "bytes_hbm": 49152,
      "arith_intensity": 21.333333333333332,
      "tile_count_expected": 8,
      "sim_wall_clock_s": 0.639,
      "engines": {
        "pe_dma": {
          "occupancy_ns": 16594.192,
          "wall_ns": 15271.192000000001,
          "record_count": 576
        },
        "pe_fetch_store": {
          "occupancy_ns": 8704.0,
          "wall_ns": 7624.0,
          "record_count": 320
        },
        "pe_gemm": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "pe_math": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "stages": {
        "DMA_READ": {
          "occupancy_ns": 15250.192000000001,
          "wall_ns": 15250.192000000001,
          "record_count": 512
        },
        "DMA_WRITE": {
          "occupancy_ns": 1344.0,
          "wall_ns": 1344.0,
          "record_count": 64
        },
        "FETCH": {
          "occupancy_ns": 8192.0,
          "wall_ns": 7616.0,
          "record_count": 256
        },
        "STORE": {
          "occupancy_ns": 512.0,
          "wall_ns": 512.0,
          "record_count": 64
        },
        "GEMM": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "MATH": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "pe_window_ns": 15327.576000000001,
      "composite_window_ns": 15327.576000000001
    },
    {
      "M": 32,
      "K": 128,
      "N": 128,
      "variant": "load_load",
      "flops": 1048576,
      "bytes_hbm": 49152,
      "arith_intensity": 21.333333333333332,
      "tile_count_expected": 8,
      "sim_wall_clock_s": 0.684,
      "engines": {
        "pe_dma": {
          "occupancy_ns": 16594.192,
          "wall_ns": 15271.192000000001,
          "record_count": 576
        },
        "pe_fetch_store": {
          "occupancy_ns": 8704.0,
          "wall_ns": 7624.0,
          "record_count": 320
        },
        "pe_gemm": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "pe_math": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "stages": {
        "DMA_READ": {
          "occupancy_ns": 15250.192000000001,
          "wall_ns": 15250.192000000001,
          "record_count": 512
        },
        "DMA_WRITE": {
          "occupancy_ns": 1344.0,
          "wall_ns": 1344.0,
          "record_count": 64
        },
        "FETCH": {
          "occupancy_ns": 8192.0,
          "wall_ns": 7616.0,
          "record_count": 256
        },
        "STORE": {
          "occupancy_ns": 512.0,
          "wall_ns": 512.0,
          "record_count": 64
        },
        "GEMM": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "MATH": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "pe_window_ns": 15327.576000000001,
      "composite_window_ns": 15327.576000000001
    },
    {
      "M": 32,
      "K": 3072,
      "N": 32,
      "variant": "ref_ref",
      "flops": 6291456,
      "bytes_hbm": 395264,
      "arith_intensity": 15.917098445595855,
      "tile_count_expected": 48,
      "sim_wall_clock_s": 0.623,
      "engines": {
        "pe_dma": {
          "occupancy_ns": 16594.192,
          "wall_ns": 15271.192000000001,
          "record_count": 576
        },
        "pe_fetch_store": {
          "occupancy_ns": 8704.0,
          "wall_ns": 7624.0,
          "record_count": 320
        },
        "pe_gemm": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "pe_math": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "stages": {
        "DMA_READ": {
          "occupancy_ns": 15250.192000000001,
          "wall_ns": 15250.192000000001,
          "record_count": 512
        },
        "DMA_WRITE": {
          "occupancy_ns": 1344.0,
          "wall_ns": 1344.0,
          "record_count": 64
        },
        "FETCH": {
          "occupancy_ns": 8192.0,
          "wall_ns": 7616.0,
          "record_count": 256
        },
        "STORE": {
          "occupancy_ns": 512.0,
          "wall_ns": 512.0,
          "record_count": 64
        },
        "GEMM": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "MATH": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "pe_window_ns": 15327.576000000001,
      "composite_window_ns": 15327.576000000001
    },
    {
      "M": 32,
      "K": 3072,
      "N": 32,
      "variant": "load_ref",
      "flops": 6291456,
      "bytes_hbm": 395264,
      "arith_intensity": 15.917098445595855,
      "tile_count_expected": 48,
      "sim_wall_clock_s": 0.763,
      "engines": {
        "pe_dma": {
          "occupancy_ns": 16594.192,
          "wall_ns": 15271.192000000001,
          "record_count": 576
        },
        "pe_fetch_store": {
          "occupancy_ns": 8704.0,
          "wall_ns": 7624.0,
          "record_count": 320
        },
        "pe_gemm": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "pe_math": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "stages": {
        "DMA_READ": {
          "occupancy_ns": 15250.192000000001,
          "wall_ns": 15250.192000000001,
          "record_count": 512
        },
        "DMA_WRITE": {
          "occupancy_ns": 1344.0,
          "wall_ns": 1344.0,
          "record_count": 64
        },
        "FETCH": {
          "occupancy_ns": 8192.0,
          "wall_ns": 7616.0,
          "record_count": 256
        },
        "STORE": {
          "occupancy_ns": 512.0,
          "wall_ns": 512.0,
          "record_count": 64
        },
        "GEMM": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "MATH": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "pe_window_ns": 15327.576000000001,
      "composite_window_ns": 15327.576000000001
    },
    {
      "M": 32,
      "K": 3072,
      "N": 32,
      "variant": "load_load",
      "flops": 6291456,
      "bytes_hbm": 395264,
      "arith_intensity": 15.917098445595855,
      "tile_count_expected": 48,
      "sim_wall_clock_s": 0.647,
      "engines": {
        "pe_dma": {
          "occupancy_ns": 16594.192,
          "wall_ns": 15271.192000000001,
          "record_count": 576
        },
        "pe_fetch_store": {
          "occupancy_ns": 8704.0,
          "wall_ns": 7624.0,
          "record_count": 320
        },
        "pe_gemm": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "pe_math": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "stages": {
        "DMA_READ": {
          "occupancy_ns": 15250.192000000001,
          "wall_ns": 15250.192000000001,
          "record_count": 512
        },
        "DMA_WRITE": {
          "occupancy_ns": 1344.0,
          "wall_ns": 1344.0,
          "record_count": 64
        },
        "FETCH": {
          "occupancy_ns": 8192.0,
          "wall_ns": 7616.0,
          "record_count": 256
        },
        "STORE": {
          "occupancy_ns": 512.0,
          "wall_ns": 512.0,
          "record_count": 64
        },
        "GEMM": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "MATH": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "pe_window_ns": 15327.576000000001,
      "composite_window_ns": 15327.576000000001
    },
    {
      "M": 8,
      "K": 128,
      "N": 128,
      "variant": "ref_ref",
      "flops": 262144,
      "bytes_hbm": 36864,
      "arith_intensity": 7.111111111111111,
      "tile_count_expected": 8,
      "sim_wall_clock_s": 0.637,
      "engines": {
        "pe_dma": {
          "occupancy_ns": 16594.192,
          "wall_ns": 15271.192000000001,
          "record_count": 576
        },
        "pe_fetch_store": {
          "occupancy_ns": 8704.0,
          "wall_ns": 7624.0,
          "record_count": 320
        },
        "pe_gemm": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "pe_math": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "stages": {
        "DMA_READ": {
          "occupancy_ns": 15250.192000000001,
          "wall_ns": 15250.192000000001,
          "record_count": 512
        },
        "DMA_WRITE": {
          "occupancy_ns": 1344.0,
          "wall_ns": 1344.0,
          "record_count": 64
        },
        "FETCH": {
          "occupancy_ns": 8192.0,
          "wall_ns": 7616.0,
          "record_count": 256
        },
        "STORE": {
          "occupancy_ns": 512.0,
          "wall_ns": 512.0,
          "record_count": 64
        },
        "GEMM": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "MATH": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "pe_window_ns": 15327.576000000001,
      "composite_window_ns": 15327.576000000001
    },
    {
      "M": 8,
      "K": 128,
      "N": 128,
      "variant": "load_ref",
      "flops": 262144,
      "bytes_hbm": 36864,
      "arith_intensity": 7.111111111111111,
      "tile_count_expected": 8,
      "sim_wall_clock_s": 0.678,
      "engines": {
        "pe_dma": {
          "occupancy_ns": 16594.192,
          "wall_ns": 15271.192000000001,
          "record_count": 576
        },
        "pe_fetch_store": {
          "occupancy_ns": 8704.0,
          "wall_ns": 7624.0,
          "record_count": 320
        },
        "pe_gemm": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "pe_math": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "stages": {
        "DMA_READ": {
          "occupancy_ns": 15250.192000000001,
          "wall_ns": 15250.192000000001,
          "record_count": 512
        },
        "DMA_WRITE": {
          "occupancy_ns": 1344.0,
          "wall_ns": 1344.0,
          "record_count": 64
        },
        "FETCH": {
          "occupancy_ns": 8192.0,
          "wall_ns": 7616.0,
          "record_count": 256
        },
        "STORE": {
          "occupancy_ns": 512.0,
          "wall_ns": 512.0,
          "record_count": 64
        },
        "GEMM": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "MATH": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "pe_window_ns": 15327.576000000001,
      "composite_window_ns": 15327.576000000001
    },
    {
      "M": 8,
      "K": 128,
      "N": 128,
      "variant": "load_load",
      "flops": 262144,
      "bytes_hbm": 36864,
      "arith_intensity": 7.111111111111111,
      "tile_count_expected": 8,
      "sim_wall_clock_s": 0.66,
      "engines": {
        "pe_dma": {
          "occupancy_ns": 16594.192,
          "wall_ns": 15271.192000000001,
          "record_count": 576
        },
        "pe_fetch_store": {
          "occupancy_ns": 8704.0,
          "wall_ns": 7624.0,
          "record_count": 320
        },
        "pe_gemm": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "pe_math": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "stages": {
        "DMA_READ": {
          "occupancy_ns": 15250.192000000001,
          "wall_ns": 15250.192000000001,
          "record_count": 512
        },
        "DMA_WRITE": {
          "occupancy_ns": 1344.0,
          "wall_ns": 1344.0,
          "record_count": 64
        },
        "FETCH": {
          "occupancy_ns": 8192.0,
          "wall_ns": 7616.0,
          "record_count": 256
        },
        "STORE": {
          "occupancy_ns": 512.0,
          "wall_ns": 512.0,
          "record_count": 64
        },
        "GEMM": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "MATH": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "pe_window_ns": 15327.576000000001,
      "composite_window_ns": 15327.576000000001
    },
    {
      "M": 128,
      "K": 8,
      "N": 128,
      "variant": "ref_ref",
      "flops": 262144,
      "bytes_hbm": 36864,
      "arith_intensity": 7.111111111111111,
      "tile_count_expected": 16,
      "sim_wall_clock_s": 0.658,
      "engines": {
        "pe_dma": {
          "occupancy_ns": 16594.192,
          "wall_ns": 15271.192000000001,
          "record_count": 576
        },
        "pe_fetch_store": {
          "occupancy_ns": 8704.0,
          "wall_ns": 7624.0,
          "record_count": 320
        },
        "pe_gemm": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "pe_math": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "stages": {
        "DMA_READ": {
          "occupancy_ns": 15250.192000000001,
          "wall_ns": 15250.192000000001,
          "record_count": 512
        },
        "DMA_WRITE": {
          "occupancy_ns": 1344.0,
          "wall_ns": 1344.0,
          "record_count": 64
        },
        "FETCH": {
          "occupancy_ns": 8192.0,
          "wall_ns": 7616.0,
          "record_count": 256
        },
        "STORE": {
          "occupancy_ns": 512.0,
          "wall_ns": 512.0,
          "record_count": 64
        },
        "GEMM": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "MATH": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "pe_window_ns": 15327.576000000001,
      "composite_window_ns": 15327.576000000001
    },
    {
      "M": 128,
      "K": 8,
      "N": 128,
      "variant": "load_ref",
      "flops": 262144,
      "bytes_hbm": 36864,
      "arith_intensity": 7.111111111111111,
      "tile_count_expected": 16,
      "sim_wall_clock_s": 0.646,
      "engines": {
        "pe_dma": {
          "occupancy_ns": 16594.192,
          "wall_ns": 15271.192000000001,
          "record_count": 576
        },
        "pe_fetch_store": {
          "occupancy_ns": 8704.0,
          "wall_ns": 7624.0,
          "record_count": 320
        },
        "pe_gemm": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "pe_math": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "stages": {
        "DMA_READ": {
          "occupancy_ns": 15250.192000000001,
          "wall_ns": 15250.192000000001,
          "record_count": 512
        },
        "DMA_WRITE": {
          "occupancy_ns": 1344.0,
          "wall_ns": 1344.0,
          "record_count": 64
        },
        "FETCH": {
          "occupancy_ns": 8192.0,
          "wall_ns": 7616.0,
          "record_count": 256
        },
        "STORE": {
          "occupancy_ns": 512.0,
          "wall_ns": 512.0,
          "record_count": 64
        },
        "GEMM": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "MATH": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "pe_window_ns": 15327.576000000001,
      "composite_window_ns": 15327.576000000001
    },
    {
      "M": 128,
      "K": 8,
      "N": 128,
      "variant": "load_load",
      "flops": 262144,
      "bytes_hbm": 36864,
      "arith_intensity": 7.111111111111111,
      "tile_count_expected": 16,
      "sim_wall_clock_s": 0.673,
      "engines": {
        "pe_dma": {
          "occupancy_ns": 16594.192,
          "wall_ns": 15271.192000000001,
          "record_count": 576
        },
        "pe_fetch_store": {
          "occupancy_ns": 8704.0,
          "wall_ns": 7624.0,
          "record_count": 320
        },
        "pe_gemm": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "pe_math": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "stages": {
        "DMA_READ": {
          "occupancy_ns": 15250.192000000001,
          "wall_ns": 15250.192000000001,
          "record_count": 512
        },
        "DMA_WRITE": {
          "occupancy_ns": 1344.0,
          "wall_ns": 1344.0,
          "record_count": 64
        },
        "FETCH": {
          "occupancy_ns": 8192.0,
          "wall_ns": 7616.0,
          "record_count": 256
        },
        "STORE": {
          "occupancy_ns": 512.0,
          "wall_ns": 512.0,
          "record_count": 64
        },
        "GEMM": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "MATH": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "pe_window_ns": 15327.576000000001,
      "composite_window_ns": 15327.576000000001
    },
    {
      "M": 512,
      "K": 512,
      "N": 512,
      "variant": "ref_ref",
      "flops": 268435456,
      "bytes_hbm": 1572864,
      "arith_intensity": 170.66666666666666,
      "tile_count_expected": 2048,
      "sim_wall_clock_s": 0.727,
      "engines": {
        "pe_dma": {
          "occupancy_ns": 16594.192,
          "wall_ns": 15271.192000000001,
          "record_count": 576
        },
        "pe_fetch_store": {
          "occupancy_ns": 8704.0,
          "wall_ns": 7624.0,
          "record_count": 320
        },
        "pe_gemm": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "pe_math": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "stages": {
        "DMA_READ": {
          "occupancy_ns": 15250.192000000001,
          "wall_ns": 15250.192000000001,
          "record_count": 512
        },
        "DMA_WRITE": {
          "occupancy_ns": 1344.0,
          "wall_ns": 1344.0,
          "record_count": 64
        },
        "FETCH": {
          "occupancy_ns": 8192.0,
          "wall_ns": 7616.0,
          "record_count": 256
        },
        "STORE": {
          "occupancy_ns": 512.0,
          "wall_ns": 512.0,
          "record_count": 64
        },
        "GEMM": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "MATH": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "pe_window_ns": 15327.576000000001,
      "composite_window_ns": 15327.576000000001
    },
    {
      "M": 512,
      "K": 512,
      "N": 512,
      "variant": "load_ref",
      "flops": 268435456,
      "bytes_hbm": 1572864,
      "arith_intensity": 170.66666666666666,
      "tile_count_expected": 2048,
      "sim_wall_clock_s": 0.707,
      "engines": {
        "pe_dma": {
          "occupancy_ns": 16594.192,
          "wall_ns": 15271.192000000001,
          "record_count": 576
        },
        "pe_fetch_store": {
          "occupancy_ns": 8704.0,
          "wall_ns": 7624.0,
          "record_count": 320
        },
        "pe_gemm": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "pe_math": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "stages": {
        "DMA_READ": {
          "occupancy_ns": 15250.192000000001,
          "wall_ns": 15250.192000000001,
          "record_count": 512
        },
        "DMA_WRITE": {
          "occupancy_ns": 1344.0,
          "wall_ns": 1344.0,
          "record_count": 64
        },
        "FETCH": {
          "occupancy_ns": 8192.0,
          "wall_ns": 7616.0,
          "record_count": 256
        },
        "STORE": {
          "occupancy_ns": 512.0,
          "wall_ns": 512.0,
          "record_count": 64
        },
        "GEMM": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "MATH": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "pe_window_ns": 15327.576000000001,
      "composite_window_ns": 15327.576000000001
    },
    {
      "M": 512,
      "K": 512,
      "N": 512,
      "variant": "load_load",
      "flops": 268435456,
      "bytes_hbm": 1572864,
      "arith_intensity": 170.66666666666666,
      "tile_count_expected": 2048,
      "sim_wall_clock_s": 0.828,
      "engines": {
        "pe_dma": {
          "occupancy_ns": 16594.192,
          "wall_ns": 15271.192000000001,
          "record_count": 576
        },
        "pe_fetch_store": {
          "occupancy_ns": 8704.0,
          "wall_ns": 7624.0,
          "record_count": 320
        },
        "pe_gemm": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "pe_math": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "stages": {
        "DMA_READ": {
          "occupancy_ns": 15250.192000000001,
          "wall_ns": 15250.192000000001,
          "record_count": 512
        },
        "DMA_WRITE": {
          "occupancy_ns": 1344.0,
          "wall_ns": 1344.0,
          "record_count": 64
        },
        "FETCH": {
          "occupancy_ns": 8192.0,
          "wall_ns": 7616.0,
          "record_count": 256
        },
        "STORE": {
          "occupancy_ns": 512.0,
          "wall_ns": 512.0,
          "record_count": 64
        },
        "GEMM": {
          "occupancy_ns": 4194.304000000004,
          "wall_ns": 4194.304000000004,
          "record_count": 256
        },
        "MATH": {
          "occupancy_ns": 0,
          "wall_ns": 0.0,
          "record_count": 0
        }
      },
      "pe_window_ns": 15327.576000000001,
      "composite_window_ns": 15327.576000000001
    }
  ]
}