{
  "clockHz": 1500000000,
  "tileCount": 1472,
  "sampleCount": 233362,
  "matchedSampleCount": 130762,
  "profileSpanCycles": 16202412,
  "profileSpanMs": 10.801608,
  "atOffset": null,
  "groups": [
    {
      "name": "repeat-remainder",
      "dimensions": {},
      "sampleCount": 1472,
      "tileCount": 1472,
      "phaseCount": 9,
      "phaseCycles": 15303324,
      "phaseMs": 10.202216,
      "workCycles": 22520862120,
      "averageActiveTiles": 1471.6320532715638,
      "usefulCycles": 0.0,
      "estimatedWorkCycles": 0,
      "usefulUtilization": null,
      "workLaneOccupancy": null,
      "deviceUsefulUtilization": null,
      "meanCycles": 15299498.722826088,
      "p50Cycles": 15299526,
      "p95Cycles": 15299598,
      "maximumCycles": 15299622,
      "firstOffset": 694152,
      "lastOffset": 15997476
    },
    {
      "name": "reduce_sum_f16",
      "dimensions": {},
      "sampleCount": 22098,
      "tileCount": 1472,
      "phaseCount": 137,
      "phaseCycles": 122358,
      "phaseMs": 0.081572,
      "workCycles": 86301906,
      "averageActiveTiles": 705.3229539547884,
      "usefulCycles": 30933608.0,
      "estimatedWorkCycles": 86301906,
      "usefulUtilization": 0.35843481834572694,
      "workLaneOccupancy": 0.9908698136754196,
      "deviceUsefulUtilization": 0.1717474897254457,
      "meanCycles": 3905.417051316861,
      "p50Cycles": 2154,
      "p95Cycles": 8748,
      "maximumCycles": 10386,
      "firstOffset": 24600,
      "lastOffset": 16197570
    },
    {
      "name": "fill_zero_u64",
      "dimensions": {},
      "sampleCount": 35760,
      "tileCount": 1472,
      "phaseCount": 183,
      "phaseCycles": 87006,
      "phaseMs": 0.058004,
      "workCycles": 11801430,
      "averageActiveTiles": 135.6392662574995,
      "usefulCycles": 0.0,
      "estimatedWorkCycles": 7339734,
      "usefulUtilization": 0.0,
      "workLaneOccupancy": null,
      "deviceUsefulUtilization": 0.0,
      "meanCycles": 330.01761744966444,
      "p50Cycles": 306,
      "p95Cycles": 726,
      "maximumCycles": 3150,
      "firstOffset": 3048,
      "lastOffset": 16190028
    },
    {
      "name": "gemm_f8_init_large_rows_interleaved_k192_c144_r180_r184",
      "dimensions": {},
      "sampleCount": 920,
      "tileCount": 920,
      "phaseCount": 22,
      "phaseCycles": 57882,
      "phaseMs": 0.038588,
      "workCycles": 44739600,
      "averageActiveTiles": 772.944956981445,
      "usefulCycles": 35637120.0,
      "estimatedWorkCycles": 44739600,
      "usefulUtilization": 0.7965453423812462,
      "workLaneOccupancy": 0.9746376811594203,
      "deviceUsefulUtilization": 0.418264745516741,
      "meanCycles": 48630.0,
      "p50Cycles": 48630,
      "p95Cycles": 48630,
      "maximumCycles": 48630,
      "firstOffset": 605178,
      "lastOffset": 663060
    },
    {
      "name": "gemm_f8_init_small_rows_interleaved_k192_c144_r180_r184",
      "dimensions": {},
      "sampleCount": 552,
      "tileCount": 552,
      "phaseCount": 21,
      "phaseCycles": 55308,
      "phaseMs": 0.036872,
      "workCycles": 26515872,
      "averageActiveTiles": 479.4220004339336,
      "usefulCycles": 20839968.0,
      "estimatedWorkCycles": 26515872,
      "usefulUtilization": 0.7859431513321531,
      "workLaneOccupancy": 0.971027911969941,
      "deviceUsefulUtilization": 0.2559771996188931,
      "meanCycles": 48036.0,
      "p50Cycles": 48036,
      "p95Cycles": 48036,
      "maximumCycles": 48036,
      "firstOffset": 607080,
      "lastOffset": 662388
    },
    {
      "name": "attention_softmax_d72_p768_tail_f8_sm4",
      "dimensions": {},
      "sampleCount": 1472,
      "tileCount": 1472,
      "phaseCount": 15,
      "phaseCycles": 53034,
      "phaseMs": 0.035356,
      "workCycles": 71468532,
      "averageActiveTiles": 1347.5983708564318,
      "usefulCycles": 33499008.0,
      "estimatedWorkCycles": 71468532,
      "usefulUtilization": 0.46872388535978327,
      "workLaneOccupancy": 0.9972222222222222,
      "deviceUsefulUtilization": 0.4291111034594706,
      "meanCycles": 48551.99184782609,
      "p50Cycles": 48552,
      "p95Cycles": 48552,
      "maximumCycles": 48552,
      "firstOffset": 270732,
      "lastOffset": 323766
    },
    {
      "name": "gemm_f8_init_large_rows_interleaved_k384_c80_r160_r164",
      "dimensions": {},
      "sampleCount": 795,
      "tileCount": 795,
      "phaseCount": 13,
      "phaseCycles": 49200,
      "phaseMs": 0.0328,
      "workCycles": 38951820,
      "averageActiveTiles": 791.7036585365854,
      "usefulCycles": 31291200.0,
      "estimatedWorkCycles": 38951820,
      "usefulUtilization": 0.8033308841538085,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.4320652173913043,
      "meanCycles": 48996.0,
      "p50Cycles": 48996,
      "p95Cycles": 48996,
      "maximumCycles": 48996,
      "firstOffset": 487710,
      "lastOffset": 536910
    },
    {
      "name": "copy_strided_u64",
      "dimensions": {},
      "sampleCount": 8536,
      "tileCount": 1472,
      "phaseCount": 98,
      "phaseCycles": 49002,
      "phaseMs": 0.032668,
      "workCycles": 10896132,
      "averageActiveTiles": 222.3609648585772,
      "usefulCycles": 0.0,
      "estimatedWorkCycles": 0,
      "usefulUtilization": null,
      "workLaneOccupancy": null,
      "deviceUsefulUtilization": null,
      "meanCycles": 1276.4915651358951,
      "p50Cycles": 984,
      "p95Cycles": 3060,
      "maximumCycles": 3060,
      "firstOffset": 23694,
      "lastOffset": 16119468
    },
    {
      "name": "gemm_f8_init_small_rows_interleaved_k384_c80_r160_r164",
      "dimensions": {},
      "sampleCount": 636,
      "tileCount": 636,
      "phaseCount": 15,
      "phaseCycles": 48462,
      "phaseMs": 0.032308,
      "workCycles": 30577608,
      "averageActiveTiles": 630.9605051380463,
      "usefulCycles": 24346080.0,
      "estimatedWorkCycles": 30577608,
      "usefulUtilization": 0.7962061649819044,
      "workLaneOccupancy": 0.996875,
      "deviceUsefulUtilization": 0.34128712231726155,
      "meanCycles": 48078.0,
      "p50Cycles": 48078,
      "p95Cycles": 48078,
      "maximumCycles": 48078,
      "firstOffset": 487434,
      "lastOffset": 535896
    },
    {
      "name": "cast_f16_f8",
      "dimensions": {},
      "sampleCount": 12514,
      "tileCount": 1472,
      "phaseCount": 130,
      "phaseCycles": 48294,
      "phaseMs": 0.032196,
      "workCycles": 39973530,
      "averageActiveTiles": 827.7121381538079,
      "usefulCycles": 1972820.0,
      "estimatedWorkCycles": 37844622,
      "usefulUtilization": 0.05212946769556848,
      "workLaneOccupancy": 0.9384120980109366,
      "deviceUsefulUtilization": 0.02775150077154242,
      "meanCycles": 3194.304778647914,
      "p50Cycles": 3048,
      "p95Cycles": 9300,
      "maximumCycles": 9300,
      "firstOffset": 0,
      "lastOffset": 16189182
    },
    {
      "name": "copy_u64",
      "dimensions": {},
      "sampleCount": 9311,
      "tileCount": 1472,
      "phaseCount": 100,
      "phaseCycles": 48048,
      "phaseMs": 0.032032,
      "workCycles": 12473154,
      "averageActiveTiles": 259.5977772227772,
      "usefulCycles": 0.0,
      "estimatedWorkCycles": 0,
      "usefulUtilization": null,
      "workLaneOccupancy": null,
      "deviceUsefulUtilization": null,
      "meanCycles": 1339.61486413919,
      "p50Cycles": 834,
      "p95Cycles": 3018,
      "maximumCycles": 10578,
      "firstOffset": 94680,
      "lastOffset": 16192212
    },
    {
      "name": "gemm_f8_init_large_rows_interleaved_k288_c96_r144_r148",
      "dimensions": {},
      "sampleCount": 720,
      "tileCount": 720,
      "phaseCount": 1,
      "phaseCycles": 40866,
      "phaseMs": 0.027244,
      "workCycles": 29350080,
      "averageActiveTiles": 718.2029070621054,
      "usefulCycles": 23016960.0,
      "estimatedWorkCycles": 29350080,
      "usefulUtilization": 0.7842213717986458,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.38262912296604595,
      "meanCycles": 40764.0,
      "p50Cycles": 40764,
      "p95Cycles": 40764,
      "maximumCycles": 40764,
      "firstOffset": 131130,
      "lastOffset": 171996
    },
    {
      "name": "gemm_f8_init_small_rows_interleaved_k288_c96_r144_r148",
      "dimensions": {},
      "sampleCount": 720,
      "tileCount": 720,
      "phaseCount": 3,
      "phaseCycles": 40416,
      "phaseMs": 0.026944,
      "workCycles": 28922400,
      "averageActiveTiles": 715.6175771971497,
      "usefulCycles": 22332672.0,
      "estimatedWorkCycles": 28922400,
      "usefulUtilization": 0.7721583271097834,
      "workLaneOccupancy": 0.9972222222222222,
      "deviceUsefulUtilization": 0.37538727667045335,
      "meanCycles": 40170.0,
      "p50Cycles": 40170,
      "p95Cycles": 40170,
      "maximumCycles": 40170,
      "firstOffset": 130920,
      "lastOffset": 171336
    },
    {
      "name": "gemm_f8_init_large_rows_interleaved_k384_c64_r160_r164",
      "dimensions": {},
      "sampleCount": 15,
      "tileCount": 15,
      "phaseCount": 10,
      "phaseCycles": 39468,
      "phaseMs": 0.026312,
      "workCycles": 589140,
      "averageActiveTiles": 14.927029492246884,
      "usefulCycles": 472320.0,
      "estimatedWorkCycles": 589140,
      "usefulUtilization": 0.8017109685304002,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.008129866352928734,
      "meanCycles": 39276.0,
      "p50Cycles": 39276,
      "p95Cycles": 39276,
      "maximumCycles": 39276,
      "firstOffset": 487716,
      "lastOffset": 527184
    },
    {
      "name": "gemm_f8_init_small_rows_interleaved_k384_c64_r160_r164",
      "dimensions": {},
      "sampleCount": 12,
      "tileCount": 12,
      "phaseCount": 11,
      "phaseCycles": 38850,
      "phaseMs": 0.0259,
      "workCycles": 462528,
      "averageActiveTiles": 11.905482625482625,
      "usefulCycles": 367488.0,
      "estimatedWorkCycles": 462528,
      "usefulUtilization": 0.7945205479452054,
      "workLaneOccupancy": 0.996875,
      "deviceUsefulUtilization": 0.006426053382575122,
      "meanCycles": 38544.0,
      "p50Cycles": 38544,
      "p95Cycles": 38544,
      "maximumCycles": 38544,
      "firstOffset": 487440,
      "lastOffset": 526290
    },
    {
      "name": "gemm_f8_init_large_rows_interleaved_k96_c576_r20_r24",
      "dimensions": {},
      "sampleCount": 1440,
      "tileCount": 1440,
      "phaseCount": 19,
      "phaseCycles": 30150,
      "phaseMs": 0.0201,
      "workCycles": 41489280,
      "averageActiveTiles": 1376.0955223880596,
      "usefulCycles": 14929920.0,
      "estimatedWorkCycles": 41489280,
      "usefulUtilization": 0.35985006247396917,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.336404931862427,
      "meanCycles": 28812.0,
      "p50Cycles": 28812,
      "p95Cycles": 28812,
      "maximumCycles": 28812,
      "firstOffset": 395382,
      "lastOffset": 425532
    },
    {
      "name": "gemm_f8_init_small_rows_interleaved_k192_c96_r144_r148",
      "dimensions": {},
      "sampleCount": 720,
      "tileCount": 720,
      "phaseCount": 6,
      "phaseCycles": 29532,
      "phaseMs": 0.019688,
      "workCycles": 19418400,
      "averageActiveTiles": 657.5375863470134,
      "usefulCycles": 14888448.0,
      "estimatedWorkCycles": 19418400,
      "usefulUtilization": 0.7667185761957731,
      "workLaneOccupancy": 0.9972222222222222,
      "deviceUsefulUtilization": 0.34249068070596966,
      "meanCycles": 26970.0,
      "p50Cycles": 26970,
      "p95Cycles": 26970,
      "maximumCycles": 26970,
      "firstOffset": 16025658,
      "lastOffset": 16055190
    },
    {
      "name": "gemm_f8_init_large_rows_interleaved_k192_c96_r144_r148",
      "dimensions": {},
      "sampleCount": 720,
      "tileCount": 720,
      "phaseCount": 4,
      "phaseCycles": 27642,
      "phaseMs": 0.018428,
      "workCycles": 19699200,
      "averageActiveTiles": 712.6546559583243,
      "usefulCycles": 15344640.0,
      "estimatedWorkCycles": 19699200,
      "usefulUtilization": 0.7789473684210526,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.3771198837308066,
      "meanCycles": 27360.0,
      "p50Cycles": 27360,
      "p95Cycles": 27360,
      "maximumCycles": 27360,
      "firstOffset": 16028046,
      "lastOffset": 16055688
    },
    {
      "name": "add_f16",
      "dimensions": {},
      "sampleCount": 11220,
      "tileCount": 1472,
      "phaseCount": 141,
      "phaseCycles": 27540,
      "phaseMs": 0.01836,
      "workCycles": 20423970,
      "averageActiveTiles": 741.6111111111111,
      "usefulCycles": 4623400.0,
      "estimatedWorkCycles": 20423970,
      "usefulUtilization": 0.22637126866128376,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.11404853809478703,
      "meanCycles": 1820.3181818181818,
      "p50Cycles": 1374,
      "p95Cycles": 4314,
      "maximumCycles": 4314,
      "firstOffset": 91452,
      "lastOffset": 16202412
    },
    {
      "name": "gemm_f8_init_small_rows_interleaved_k96_c576_r20_r24",
      "dimensions": {},
      "sampleCount": 24,
      "tileCount": 24,
      "phaseCount": 2,
      "phaseCycles": 27150,
      "phaseMs": 0.0181,
      "workCycles": 637056,
      "averageActiveTiles": 23.464309392265193,
      "usefulCycles": 186624.0,
      "estimatedWorkCycles": 637056,
      "usefulUtilization": 0.29294755877034356,
      "workLaneOccupancy": 0.9,
      "deviceUsefulUtilization": 0.004669709344222916,
      "meanCycles": 26544.0,
      "p50Cycles": 26544,
      "p95Cycles": 26544,
      "maximumCycles": 26544,
      "firstOffset": 395610,
      "lastOffset": 422760
    },
    {
      "name": "gemm_f16_init_large_rows_interleaved_k80_c112_r121_r122",
      "dimensions": {},
      "sampleCount": 384,
      "tileCount": 384,
      "phaseCount": 4,
      "phaseCycles": 23310,
      "phaseMs": 0.01554,
      "workCycles": 8831232,
      "averageActiveTiles": 378.86023166023165,
      "usefulCycles": 5902848.0,
      "estimatedWorkCycles": 8831232,
      "usefulUtilization": 0.6684059483433342,
      "workLaneOccupancy": 0.9,
      "deviceUsefulUtilization": 0.17203290246768507,
      "meanCycles": 22998.0,
      "p50Cycles": 22998,
      "p95Cycles": 22998,
      "maximumCycles": 22998,
      "firstOffset": 240222,
      "lastOffset": 263532
    },
    {
      "name": "gemm_f16_init_small_rows_interleaved_k80_c112_r121_r122",
      "dimensions": {},
      "sampleCount": 384,
      "tileCount": 384,
      "phaseCount": 4,
      "phaseCycles": 23232,
      "phaseMs": 0.015488,
      "workCycles": 8815104,
      "averageActiveTiles": 379.43801652892563,
      "usefulCycles": 5854464.0,
      "estimatedWorkCycles": 8815104,
      "usefulUtilization": 0.6641400940930475,
      "workLaneOccupancy": 0.9,
      "deviceUsefulUtilization": 0.17119565217391305,
      "meanCycles": 22956.0,
      "p50Cycles": 22956,
      "p95Cycles": 22956,
      "maximumCycles": 22956,
      "firstOffset": 240240,
      "lastOffset": 263472
    },
    {
      "name": "layer_norm_f16",
      "dimensions": {},
      "sampleCount": 4374,
      "tileCount": 1458,
      "phaseCount": 36,
      "phaseCycles": 22866,
      "phaseMs": 0.015244,
      "workCycles": 29463264,
      "averageActiveTiles": 1288.5184990816058,
      "usefulCycles": 11967264.0,
      "estimatedWorkCycles": 29463264,
      "usefulUtilization": 0.40617577197149646,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.35554687232610405,
      "meanCycles": 6736.0,
      "p50Cycles": 6738,
      "p95Cycles": 6738,
      "maximumCycles": 6738,
      "firstOffset": 99936,
      "lastOffset": 16007208
    },
    {
      "name": "gemm_f16_init_large_rows_interleaved_k80_c96_r121_r122",
      "dimensions": {},
      "sampleCount": 288,
      "tileCount": 288,
      "phaseCount": 3,
      "phaseCycles": 20574,
      "phaseMs": 0.013716,
      "workCycles": 5693760,
      "averageActiveTiles": 276.745406824147,
      "usefulCycles": 3702456.0,
      "estimatedWorkCycles": 5693760,
      "usefulUtilization": 0.6502655538694992,
      "workLaneOccupancy": 0.878125,
      "deviceUsefulUtilization": 0.12225407965308684,
      "meanCycles": 19770.0,
      "p50Cycles": 19770,
      "p95Cycles": 19770,
      "maximumCycles": 19770,
      "firstOffset": 239652,
      "lastOffset": 260226
    },
    {
      "name": "gemm_f16_init_small_rows_interleaved_k80_c96_r121_r122",
      "dimensions": {},
      "sampleCount": 288,
      "tileCount": 288,
      "phaseCount": 5,
      "phaseCycles": 20022,
      "phaseMs": 0.013348,
      "workCycles": 5683392,
      "averageActiveTiles": 283.8573569074019,
      "usefulCycles": 3672108.0,
      "estimatedWorkCycles": 5683392,
      "usefulUtilization": 0.6461120401337793,
      "workLaneOccupancy": 0.878125,
      "deviceUsefulUtilization": 0.1245948749853422,
      "meanCycles": 19734.0,
      "p50Cycles": 19734,
      "p95Cycles": 19734,
      "maximumCycles": 19734,
      "firstOffset": 239640,
      "lastOffset": 259662
    },
    {
      "name": "bias_gelu_f16",
      "dimensions": {},
      "sampleCount": 1458,
      "tileCount": 1458,
      "phaseCount": 31,
      "phaseCycles": 17220,
      "phaseMs": 0.01148,
      "workCycles": 24284448,
      "averageActiveTiles": 1410.2466898954704,
      "usefulCycles": 17256888.0,
      "estimatedWorkCycles": 24284448,
      "usefulUtilization": 0.7106147934678194,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.6808030980154522,
      "meanCycles": 16656.0,
      "p50Cycles": 16656,
      "p95Cycles": 16656,
      "maximumCycles": 16656,
      "firstOffset": 563364,
      "lastOffset": 580584
    },
    {
      "name": "gemm_f8_init_small_rows_interleaved_k128_c80_r104_r105",
      "dimensions": {},
      "sampleCount": 1152,
      "tileCount": 1152,
      "phaseCount": 14,
      "phaseCycles": 15006,
      "phaseMs": 0.010004,
      "workCycles": 13727232,
      "averageActiveTiles": 914.7828868452619,
      "usefulCycles": 8188128.0,
      "estimatedWorkCycles": 13727232,
      "usefulUtilization": 0.596487915407855,
      "workLaneOccupancy": 0.854296875,
      "deviceUsefulUtilization": 0.37069085409314534,
      "meanCycles": 11916.0,
      "p50Cycles": 11916,
      "p95Cycles": 11916,
      "maximumCycles": 11916,
      "firstOffset": 331536,
      "lastOffset": 346542
    },
    {
      "name": "gemm_f8_init_large_rows_interleaved_k128_c80_r104_r105",
      "dimensions": {},
      "sampleCount": 192,
      "tileCount": 192,
      "phaseCount": 6,
      "phaseCycles": 14820,
      "phaseMs": 0.00988,
      "workCycles": 2297088,
      "averageActiveTiles": 154.99919028340082,
      "usefulCycles": 1377810.0,
      "estimatedWorkCycles": 2297088,
      "usefulUtilization": 0.5998072342026078,
      "workLaneOccupancy": 0.854296875,
      "deviceUsefulUtilization": 0.06315871985565921,
      "meanCycles": 11964.0,
      "p50Cycles": 11964,
      "p95Cycles": 11964,
      "maximumCycles": 11964,
      "firstOffset": 331764,
      "lastOffset": 346584
    },
    {
      "name": "gemm_f8_init_large_rows_interleaved_k64_c48_r240_r244",
      "dimensions": {},
      "sampleCount": 1200,
      "tileCount": 1200,
      "phaseCount": 4,
      "phaseCycles": 14244,
      "phaseMs": 0.009496,
      "workCycles": 8704800,
      "averageActiveTiles": 611.1204717775905,
      "usefulCycles": 6456240.0,
      "estimatedWorkCycles": 8704800,
      "usefulUtilization": 0.7416873449131514,
      "workLaneOccupancy": 0.91875,
      "deviceUsefulUtilization": 0.30792141313505,
      "meanCycles": 7254.0,
      "p50Cycles": 7254,
      "p95Cycles": 7254,
      "maximumCycles": 7254,
      "firstOffset": 8508,
      "lastOffset": 22752
    },
    {
      "name": "rearrange_row_major_to_amp_f16_o2_r64_p64_c32_p32_b64x16",
      "dimensions": {},
      "sampleCount": 704,
      "tileCount": 704,
      "phaseCount": 6,
      "phaseCycles": 13056,
      "phaseMs": 0.008704,
      "workCycles": 7569408,
      "averageActiveTiles": 579.7647058823529,
      "usefulCycles": 720896.0,
      "estimatedWorkCycles": 7569408,
      "usefulUtilization": 0.09523809523809523,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.03751065643648764,
      "meanCycles": 10752.0,
      "p50Cycles": 10752,
      "p95Cycles": 10752,
      "maximumCycles": 10752,
      "firstOffset": 210948,
      "lastOffset": 224004
    }
  ],
  "samples": []
}
