{
  "clockHz": 1500000000,
  "tileCount": 1472,
  "sampleCount": 24023,
  "matchedSampleCount": 18135,
  "profileSpanCycles": 175578,
  "profileSpanMs": 0.117052,
  "atOffset": null,
  "groups": [
    {
      "name": "ipu_stack_gemm_f16_init_small_rows_interleaved_k160_c64_r242_r244",
      "dimensions": {},
      "sampleCount": 486,
      "tileCount": 486,
      "phaseCount": 4,
      "phaseCycles": 54114,
      "phaseMs": 0.036076,
      "workCycles": 22074120,
      "averageActiveTiles": 407.91883800864844,
      "usefulCycles": 18670752.0,
      "estimatedWorkCycles": 22074120,
      "usefulUtilization": 0.845820897956521,
      "workLaneOccupancy": 0.9921793694520967,
      "deviceUsefulUtilization": 0.23439285180560845,
      "meanCycles": 45420.0,
      "p50Cycles": 45420,
      "p95Cycles": 45420,
      "maximumCycles": 45420,
      "firstOffset": 97014,
      "lastOffset": 151128
    },
    {
      "name": "ipu_stack_gemm_f16_init_large_rows_interleaved_k160_c64_r242_r244",
      "dimensions": {},
      "sampleCount": 972,
      "tileCount": 972,
      "phaseCount": 8,
      "phaseCycles": 54036,
      "phaseMs": 0.036024,
      "workCycles": 44480664,
      "averageActiveTiles": 823.1672218520986,
      "usefulCycles": 37806336.0,
      "estimatedWorkCycles": 44480664,
      "usefulUtilization": 0.8499499018270051,
      "workLaneOccupancy": 0.9962962962962963,
      "deviceUsefulUtilization": 0.47530631752744545,
      "meanCycles": 45762.0,
      "p50Cycles": 45762,
      "p95Cycles": 45762,
      "maximumCycles": 45762,
      "firstOffset": 96918,
      "lastOffset": 150954
    },
    {
      "name": "ipu_stack_gemm_f16_init_large_rows_interleaved_k144_c48_r364_r366",
      "dimensions": {},
      "sampleCount": 680,
      "tileCount": 680,
      "phaseCount": 1,
      "phaseCycles": 44364,
      "phaseMs": 0.029576,
      "workCycles": 30114480,
      "averageActiveTiles": 678.8044360292129,
      "usefulCycles": 26879040.0,
      "estimatedWorkCycles": 30114480,
      "usefulUtilization": 0.8925619834710744,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.41159988709999884,
      "meanCycles": 44286.0,
      "p50Cycles": 44286,
      "p95Cycles": 44286,
      "maximumCycles": 44286,
      "firstOffset": 3816,
      "lastOffset": 48180
    },
    {
      "name": "ipu_stack_gemm_f16_init_small_rows_interleaved_k144_c48_r364_r366",
      "dimensions": {},
      "sampleCount": 680,
      "tileCount": 680,
      "phaseCount": 1,
      "phaseCycles": 44046,
      "phaseMs": 0.029364,
      "workCycles": 29890080,
      "averageActiveTiles": 678.6105435226808,
      "usefulCycles": 26658720.0,
      "estimatedWorkCycles": 29890080,
      "usefulUtilization": 0.8918918918918919,
      "workLaneOccupancy": 0.9972527472527473,
      "deviceUsefulUtilization": 0.4111733977718946,
      "meanCycles": 43956.0,
      "p50Cycles": 43956,
      "p95Cycles": 43956,
      "maximumCycles": 43956,
      "firstOffset": 3894,
      "lastOffset": 47940
    },
    {
      "name": "ipu_stack_copy_u64",
      "dimensions": {},
      "sampleCount": 5815,
      "tileCount": 1472,
      "phaseCount": 25,
      "phaseCycles": 32154,
      "phaseMs": 0.021436,
      "workCycles": 6624006,
      "averageActiveTiles": 206.0087702929651,
      "usefulCycles": 0.0,
      "estimatedWorkCycles": 0,
      "usefulUtilization": null,
      "workLaneOccupancy": null,
      "deviceUsefulUtilization": null,
      "meanCycles": 1139.1239896818572,
      "p50Cycles": 1050,
      "p95Cycles": 1674,
      "maximumCycles": 1674,
      "firstOffset": 53670,
      "lastOffset": 175578
    },
    {
      "name": "ipu_stack_gemm_f16_init_large_rows_interleaved_k144_c32_r364_r366",
      "dimensions": {},
      "sampleCount": 56,
      "tileCount": 56,
      "phaseCount": 1,
      "phaseCycles": 29670,
      "phaseMs": 0.01978,
      "workCycles": 1658496,
      "averageActiveTiles": 55.89807886754297,
      "usefulCycles": 1475712.0,
      "estimatedWorkCycles": 1658496,
      "usefulUtilization": 0.8897893030794165,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.03378907108629709,
      "meanCycles": 29616.0,
      "p50Cycles": 29616,
      "p95Cycles": 29616,
      "maximumCycles": 29616,
      "firstOffset": 2664,
      "lastOffset": 32334
    },
    {
      "name": "ipu_stack_gemm_f16_init_small_rows_interleaved_k144_c32_r364_r366",
      "dimensions": {},
      "sampleCount": 56,
      "tileCount": 56,
      "phaseCount": 1,
      "phaseCycles": 29442,
      "phaseMs": 0.019628,
      "workCycles": 1646736,
      "averageActiveTiles": 55.93152639087018,
      "usefulCycles": 1463616.0,
      "estimatedWorkCycles": 1646736,
      "usefulUtilization": 0.8887982044480718,
      "workLaneOccupancy": 0.9972527472527473,
      "deviceUsefulUtilization": 0.033771630589840604,
      "meanCycles": 29406.0,
      "p50Cycles": 29406,
      "p95Cycles": 29406,
      "maximumCycles": 29406,
      "firstOffset": 2742,
      "lastOffset": 32184
    },
    {
      "name": "ipu_stack_reduce_sum_f16",
      "dimensions": {},
      "sampleCount": 2930,
      "tileCount": 1472,
      "phaseCount": 18,
      "phaseCycles": 23586,
      "phaseMs": 0.015724,
      "workCycles": 17575524,
      "averageActiveTiles": 745.1676418214195,
      "usefulCycles": 10964600.0,
      "estimatedWorkCycles": 17575524,
      "usefulUtilization": 0.6238562218685486,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.315813498427586,
      "meanCycles": 5998.472354948805,
      "p50Cycles": 5574,
      "p95Cycles": 7182,
      "maximumCycles": 7368,
      "firstOffset": 54816,
      "lastOffset": 173766
    },
    {
      "name": "ipu_stack_gelu_tanh_approx_f16",
      "dimensions": {},
      "sampleCount": 1472,
      "tileCount": 1472,
      "phaseCount": 4,
      "phaseCycles": 20646,
      "phaseMs": 0.013764,
      "workCycles": 17211720,
      "averageActiveTiles": 833.658820110433,
      "usefulCycles": 9412848.0,
      "estimatedWorkCycles": 17211720,
      "usefulUtilization": 0.5468859591022861,
      "workLaneOccupancy": 0.9986301369863013,
      "deviceUsefulUtilization": 0.30972574959250976,
      "meanCycles": 11692.744565217392,
      "p50Cycles": 12060,
      "p95Cycles": 12060,
      "maximumCycles": 12060,
      "firstOffset": 60732,
      "lastOffset": 81378
    },
    {
      "name": "ipu_stack_fill_zero_u64",
      "dimensions": {},
      "sampleCount": 4842,
      "tileCount": 486,
      "phaseCount": 11,
      "phaseCycles": 8346,
      "phaseMs": 0.005564,
      "workCycles": 1179468,
      "averageActiveTiles": 141.3213515456506,
      "usefulCycles": 0.0,
      "estimatedWorkCycles": 1179468,
      "usefulUtilization": 0.0,
      "workLaneOccupancy": null,
      "deviceUsefulUtilization": 0.0,
      "meanCycles": 243.5910780669145,
      "p50Cycles": 240,
      "p95Cycles": 240,
      "maximumCycles": 1206,
      "firstOffset": 68712,
      "lastOffset": 82068
    },
    {
      "name": "ipu_stack_copy_strided_u64",
      "dimensions": {},
      "sampleCount": 146,
      "tileCount": 128,
      "phaseCount": 9,
      "phaseCycles": 6996,
      "phaseMs": 0.004664,
      "workCycles": 100596,
      "averageActiveTiles": 14.379073756432247,
      "usefulCycles": 0.0,
      "estimatedWorkCycles": 0,
      "usefulUtilization": null,
      "workLaneOccupancy": null,
      "deviceUsefulUtilization": null,
      "meanCycles": 689.013698630137,
      "p50Cycles": 618,
      "p95Cycles": 906,
      "maximumCycles": 906,
      "firstOffset": 61578,
      "lastOffset": 174672
    }
  ],
  "samples": []
}
