{
  "clockHz": 1500000000,
  "tileCount": 1472,
  "sampleCount": 42778,
  "matchedSampleCount": 26586,
  "profileSpanCycles": 276624,
  "profileSpanMs": 0.184416,
  "atOffset": null,
  "groups": [
    {
      "name": "ipu_stack_gemm_f16_init_large_rows_interleaved_k384_c16_r114_r116",
      "dimensions": {},
      "sampleCount": 2484,
      "tileCount": 828,
      "phaseCount": 6,
      "phaseCycles": 55476,
      "phaseMs": 0.036984,
      "workCycles": 37856160,
      "averageActiveTiles": 682.3880597014926,
      "usefulCycles": 27398736.0,
      "estimatedWorkCycles": 37856160,
      "usefulUtilization": 0.7237589866484081,
      "workLaneOccupancy": 0.9904891304347826,
      "deviceUsefulUtilization": 0.33551935502073754,
      "meanCycles": 15240.0,
      "p50Cycles": 15240,
      "p95Cycles": 15240,
      "maximumCycles": 15240,
      "firstOffset": 306,
      "lastOffset": 73218
    },
    {
      "name": "ipu_stack_gemm_f16_init_small_rows_interleaved_k384_c16_r114_r116",
      "dimensions": {},
      "sampleCount": 1656,
      "tileCount": 552,
      "phaseCount": 3,
      "phaseCycles": 55206,
      "phaseMs": 0.036804,
      "workCycles": 25108272,
      "averageActiveTiles": 454.8105640691229,
      "usefulCycles": 17950896.0,
      "estimatedWorkCycles": 25108272,
      "usefulUtilization": 0.7149395227198431,
      "workLaneOccupancy": 0.9904891304347826,
      "deviceUsefulUtilization": 0.2208981301654357,
      "meanCycles": 15162.0,
      "p50Cycles": 15162,
      "p95Cycles": 15162,
      "maximumCycles": 15162,
      "firstOffset": 312,
      "lastOffset": 73134
    },
    {
      "name": "ipu_stack_gemm_f16_init_large_rows_k768_c80_r7_r8",
      "dimensions": {},
      "sampleCount": 1360,
      "tileCount": 1360,
      "phaseCount": 15,
      "phaseCycles": 53796,
      "phaseMs": 0.035864,
      "workCycles": 73048320,
      "averageActiveTiles": 1357.8764220388132,
      "usefulCycles": 8922960.0,
      "estimatedWorkCycles": 73048320,
      "usefulUtilization": 0.1221514745308311,
      "workLaneOccupancy": 0.854296875,
      "deviceUsefulUtilization": 0.11268111900997974,
      "meanCycles": 53712.0,
      "p50Cycles": 53712,
      "p95Cycles": 53712,
      "maximumCycles": 53712,
      "firstOffset": 220362,
      "lastOffset": 274158
    },
    {
      "name": "ipu_stack_gemm_f16_init_small_rows_k768_c80_r7_r8",
      "dimensions": {},
      "sampleCount": 112,
      "tileCount": 112,
      "phaseCount": 3,
      "phaseCycles": 53268,
      "phaseMs": 0.035512,
      "workCycles": 5962656,
      "averageActiveTiles": 111.93692273034468,
      "usefulCycles": 642978.0,
      "estimatedWorkCycles": 5962656,
      "usefulUtilization": 0.10783415981066155,
      "workLaneOccupancy": 0.854296875,
      "deviceUsefulUtilization": 0.00820015218370765,
      "meanCycles": 53238.0,
      "p50Cycles": 53238,
      "p95Cycles": 53238,
      "maximumCycles": 53238,
      "firstOffset": 220368,
      "lastOffset": 273636
    },
    {
      "name": "ipu_stack_attention_softmax_d72_p768_tail_f16",
      "dimensions": {},
      "sampleCount": 1472,
      "tileCount": 1472,
      "phaseCount": 16,
      "phaseCycles": 42666,
      "phaseMs": 0.028444,
      "workCycles": 59359872,
      "averageActiveTiles": 1391.2687385740403,
      "usefulCycles": 34070544.0,
      "estimatedWorkCycles": 59359872,
      "usefulUtilization": 0.573965927689332,
      "workLaneOccupancy": 0.9986324786324786,
      "deviceUsefulUtilization": 0.5424869919842498,
      "meanCycles": 40326.0,
      "p50Cycles": 40326,
      "p95Cycles": 40326,
      "maximumCycles": 40326,
      "firstOffset": 146628,
      "lastOffset": 189294
    },
    {
      "name": "ipu_stack_unpack_amp_to_row_major_f16_o1_r9_p16_c116_p116",
      "dimensions": {},
      "sampleCount": 12,
      "tileCount": 6,
      "phaseCount": 4,
      "phaseCycles": 27096,
      "phaseMs": 0.018064,
      "workCycles": 147456,
      "averageActiveTiles": 5.441984056687334,
      "usefulCycles": 6264.0,
      "estimatedWorkCycles": 147456,
      "usefulUtilization": 0.04248046875,
      "workLaneOccupancy": 0.5625,
      "deviceUsefulUtilization": 0.00015705029460469057,
      "meanCycles": 12288.0,
      "p50Cycles": 12288,
      "p95Cycles": 12288,
      "maximumCycles": 12288,
      "firstOffset": 74424,
      "lastOffset": 105120
    },
    {
      "name": "ipu_stack_unpack_amp_to_row_major_f16_o1_r9_p16_c114_p114",
      "dimensions": {},
      "sampleCount": 8,
      "tileCount": 4,
      "phaseCount": 2,
      "phaseCycles": 26382,
      "phaseMs": 0.017588,
      "workCycles": 96672,
      "averageActiveTiles": 3.6643165794860133,
      "usefulCycles": 4104.0,
      "estimatedWorkCycles": 96672,
      "usefulUtilization": 0.04245283018867924,
      "workLaneOccupancy": 0.5625,
      "deviceUsefulUtilization": 0.00010567976189299028,
      "meanCycles": 12084.0,
      "p50Cycles": 12084,
      "p95Cycles": 12084,
      "maximumCycles": 12084,
      "firstOffset": 74454,
      "lastOffset": 104814
    },
    {
      "name": "ipu_stack_unpack_amp_to_row_major_f16_o1_r16_p16_c116_p116",
      "dimensions": {},
      "sampleCount": 540,
      "tileCount": 270,
      "phaseCount": 5,
      "phaseCycles": 24360,
      "phaseMs": 0.01624,
      "workCycles": 5883840,
      "averageActiveTiles": 241.5369458128079,
      "usefulCycles": 501120.0,
      "estimatedWorkCycles": 5883840,
      "usefulUtilization": 0.08516886930983847,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.013975155279503106,
      "meanCycles": 10896.0,
      "p50Cycles": 10896,
      "p95Cycles": 10896,
      "maximumCycles": 10896,
      "firstOffset": 74418,
      "lastOffset": 103764
    },
    {
      "name": "ipu_stack_unpack_amp_to_row_major_f16_o1_r16_p16_c114_p114",
      "dimensions": {},
      "sampleCount": 360,
      "tileCount": 180,
      "phaseCount": 3,
      "phaseCycles": 23748,
      "phaseMs": 0.015832,
      "workCycles": 3857760,
      "averageActiveTiles": 162.44567963617988,
      "usefulCycles": 328320.0,
      "estimatedWorkCycles": 3857760,
      "usefulUtilization": 0.0851063829787234,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.009392095261111233,
      "meanCycles": 10716.0,
      "p50Cycles": 10716,
      "p95Cycles": 10716,
      "maximumCycles": 10716,
      "firstOffset": 74430,
      "lastOffset": 103626
    },
    {
      "name": "ipu_stack_fill_zero_u64",
      "dimensions": {},
      "sampleCount": 9040,
      "tileCount": 1472,
      "phaseCount": 29,
      "phaseCycles": 21732,
      "phaseMs": 0.014488,
      "workCycles": 3219936,
      "averageActiveTiles": 148.16565433462176,
      "usefulCycles": 0.0,
      "estimatedWorkCycles": 1612800,
      "usefulUtilization": 0.0,
      "workLaneOccupancy": null,
      "deviceUsefulUtilization": 0.0,
      "meanCycles": 356.187610619469,
      "p50Cycles": 300,
      "p95Cycles": 600,
      "maximumCycles": 3276,
      "firstOffset": 91896,
      "lastOffset": 142266
    },
    {
      "name": "ipu_stack_gemm_f16_init_large_rows_interleaved_k80_c80_r91_r92",
      "dimensions": {},
      "sampleCount": 32,
      "tileCount": 32,
      "phaseCount": 3,
      "phaseCycles": 14004,
      "phaseMs": 0.009336,
      "workCycles": 434304,
      "averageActiveTiles": 31.01285347043702,
      "usefulCycles": 264960.0,
      "estimatedWorkCycles": 434304,
      "usefulUtilization": 0.610079575596817,
      "workLaneOccupancy": 0.9,
      "deviceUsefulUtilization": 0.012853470437017995,
      "meanCycles": 13572.0,
      "p50Cycles": 13572,
      "p95Cycles": 13572,
      "maximumCycles": 13572,
      "firstOffset": 127962,
      "lastOffset": 141966
    },
    {
      "name": "ipu_stack_gemm_f16_init_small_rows_interleaved_k80_c80_r91_r92",
      "dimensions": {},
      "sampleCount": 224,
      "tileCount": 224,
      "phaseCount": 5,
      "phaseCycles": 13992,
      "phaseMs": 0.009328,
      "workCycles": 3033408,
      "averageActiveTiles": 216.79588336192108,
      "usefulCycles": 1834560.0,
      "estimatedWorkCycles": 3033408,
      "usefulUtilization": 0.6047851129818342,
      "workLaneOccupancy": 0.9,
      "deviceUsefulUtilization": 0.08907263778059513,
      "meanCycles": 13542.0,
      "p50Cycles": 13542,
      "p95Cycles": 13542,
      "maximumCycles": 13542,
      "firstOffset": 127956,
      "lastOffset": 141948
    },
    {
      "name": "ipu_stack_copy_u64",
      "dimensions": {},
      "sampleCount": 2797,
      "tileCount": 1328,
      "phaseCount": 21,
      "phaseCycles": 13578,
      "phaseMs": 0.009052,
      "workCycles": 3084738,
      "averageActiveTiles": 227.1864781263809,
      "usefulCycles": 0.0,
      "estimatedWorkCycles": 0,
      "usefulUtilization": null,
      "workLaneOccupancy": null,
      "deviceUsefulUtilization": null,
      "meanCycles": 1102.8737933500179,
      "p50Cycles": 1158,
      "p95Cycles": 1182,
      "maximumCycles": 1626,
      "firstOffset": 21036,
      "lastOffset": 148968
    },
    {
      "name": "ipu_stack_reduce_sum_f16",
      "dimensions": {},
      "sampleCount": 1380,
      "tileCount": 920,
      "phaseCount": 5,
      "phaseCycles": 12450,
      "phaseMs": 0.0083,
      "workCycles": 5183280,
      "averageActiveTiles": 416.3277108433735,
      "usefulCycles": 1271808.0,
      "estimatedWorkCycles": 5183280,
      "usefulUtilization": 0.2453674121405751,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.06939759036144579,
      "meanCycles": 3756.0,
      "p50Cycles": 3792,
      "p95Cycles": 3792,
      "maximumCycles": 3792,
      "firstOffset": 21918,
      "lastOffset": 80448
    },
    {
      "name": "ipu_stack_gemm_f16_init_small_rows_interleaved_k80_c64_r91_r92",
      "dimensions": {},
      "sampleCount": 1008,
      "tileCount": 1008,
      "phaseCount": 8,
      "phaseCycles": 11850,
      "phaseMs": 0.0079,
      "workCycles": 10983168,
      "averageActiveTiles": 926.8496202531645,
      "usefulCycles": 6524154.0,
      "estimatedWorkCycles": 10983168,
      "usefulUtilization": 0.5940138582966226,
      "workLaneOccupancy": 0.8890625,
      "deviceUsefulUtilization": 0.37402277105118326,
      "meanCycles": 10896.0,
      "p50Cycles": 10896,
      "p95Cycles": 10896,
      "maximumCycles": 10896,
      "firstOffset": 127362,
      "lastOffset": 139212
    },
    {
      "name": "ipu_stack_gemm_f16_init_large_rows_interleaved_k80_c64_r91_r92",
      "dimensions": {},
      "sampleCount": 144,
      "tileCount": 144,
      "phaseCount": 4,
      "phaseCycles": 11742,
      "phaseMs": 0.007828,
      "workCycles": 1572480,
      "averageActiveTiles": 133.91926417986716,
      "usefulCycles": 942264.0,
      "estimatedWorkCycles": 1572480,
      "usefulUtilization": 0.5992216117216117,
      "workLaneOccupancy": 0.8890625,
      "deviceUsefulUtilization": 0.05451584057230455,
      "meanCycles": 10920.0,
      "p50Cycles": 10920,
      "p95Cycles": 10920,
      "maximumCycles": 10920,
      "firstOffset": 127488,
      "lastOffset": 139230
    },
    {
      "name": "ipu_stack_rearrange_row_major_to_amp_f16_o1_r64_p64_c16_p16",
      "dimensions": {},
      "sampleCount": 880,
      "tileCount": 880,
      "phaseCount": 9,
      "phaseCycles": 9342,
      "phaseMs": 0.006228,
      "workCycles": 1320000,
      "averageActiveTiles": 141.29736673089275,
      "usefulCycles": 450560.0,
      "estimatedWorkCycles": 1320000,
      "usefulUtilization": 0.3413333333333333,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.03276460677817803,
      "meanCycles": 1500.0,
      "p50Cycles": 1500,
      "p95Cycles": 1500,
      "maximumCycles": 1500,
      "firstOffset": 113850,
      "lastOffset": 123192
    },
    {
      "name": "ipu_stack_rearrange_row_major_to_amp_f16_o1_r0_p64_c0_p16",
      "dimensions": {},
      "sampleCount": 80,
      "tileCount": 80,
      "phaseCount": 3,
      "phaseCycles": 7032,
      "phaseMs": 0.004688,
      "workCycles": 216960,
      "averageActiveTiles": 30.853242320819113,
      "usefulCycles": 16000.0,
      "estimatedWorkCycles": 216960,
      "usefulUtilization": 0.07374631268436578,
      "workLaneOccupancy": 0.390625,
      "deviceUsefulUtilization": 0.00154572884206361,
      "meanCycles": 2712.0,
      "p50Cycles": 2712,
      "p95Cycles": 2712,
      "maximumCycles": 2712,
      "firstOffset": 113052,
      "lastOffset": 120084
    },
    {
      "name": "ipu_stack_rearrange_row_major_to_amp_f16_o0_r7_p7_c72_p80",
      "dimensions": {},
      "sampleCount": 112,
      "tileCount": 112,
      "phaseCount": 2,
      "phaseCycles": 3468,
      "phaseMs": 0.002312,
      "workCycles": 264768,
      "averageActiveTiles": 76.34602076124567,
      "usefulCycles": 28224.0,
      "estimatedWorkCycles": 264768,
      "usefulUtilization": 0.1065989847715736,
      "workLaneOccupancy": 0.9,
      "deviceUsefulUtilization": 0.005528809989468933,
      "meanCycles": 2364.0,
      "p50Cycles": 2364,
      "p95Cycles": 2364,
      "maximumCycles": 2364,
      "firstOffset": 89382,
      "lastOffset": 92850
    },
    {
      "name": "ipu_stack_rearrange_row_major_to_amp_f16_o0_r8_p8_c72_p80",
      "dimensions": {},
      "sampleCount": 1360,
      "tileCount": 1360,
      "phaseCount": 7,
      "phaseCycles": 3378,
      "phaseMs": 0.002252,
      "workCycles": 3215040,
      "averageActiveTiles": 951.7584369449378,
      "usefulCycles": 391680.0,
      "estimatedWorkCycles": 3215040,
      "usefulUtilization": 0.1218274111675127,
      "workLaneOccupancy": 0.9,
      "deviceUsefulUtilization": 0.07877056143331532,
      "meanCycles": 2364.0,
      "p50Cycles": 2364,
      "p95Cycles": 2364,
      "maximumCycles": 2364,
      "firstOffset": 89532,
      "lastOffset": 92910
    },
    {
      "name": "ipu_stack_attention_merge_v72_p80_k768_f16",
      "dimensions": {},
      "sampleCount": 1472,
      "tileCount": 1472,
      "phaseCount": 16,
      "phaseCycles": 3018,
      "phaseMs": 0.002012,
      "workCycles": 3629952,
      "averageActiveTiles": 1202.7673956262424,
      "usefulCycles": 0.0,
      "estimatedWorkCycles": 0,
      "usefulUtilization": null,
      "workLaneOccupancy": null,
      "deviceUsefulUtilization": null,
      "meanCycles": 2466.0,
      "p50Cycles": 2466,
      "p95Cycles": 2466,
      "maximumCycles": 2466,
      "firstOffset": 273606,
      "lastOffset": 276624
    },
    {
      "name": "ipu_stack_copy_strided_u64",
      "dimensions": {},
      "sampleCount": 52,
      "tileCount": 52,
      "phaseCount": 5,
      "phaseCycles": 2676,
      "phaseMs": 0.001784,
      "workCycles": 33000,
      "averageActiveTiles": 12.331838565022421,
      "usefulCycles": 0.0,
      "estimatedWorkCycles": 0,
      "usefulUtilization": null,
      "workLaneOccupancy": null,
      "deviceUsefulUtilization": null,
      "meanCycles": 634.6153846153846,
      "p50Cycles": 618,
      "p95Cycles": 834,
      "maximumCycles": 834,
      "firstOffset": 104052,
      "lastOffset": 148638
    },
    {
      "name": "ipu_stack_copy_strided_u32",
      "dimensions": {},
      "sampleCount": 1,
      "tileCount": 1,
      "phaseCount": 1,
      "phaseCycles": 1086,
      "phaseMs": 0.000724,
      "workCycles": 1086,
      "averageActiveTiles": 1.0,
      "usefulCycles": 0.0,
      "estimatedWorkCycles": 0,
      "usefulUtilization": null,
      "workLaneOccupancy": null,
      "deviceUsefulUtilization": null,
      "meanCycles": 1086.0,
      "p50Cycles": 1086,
      "p95Cycles": 1086,
      "maximumCycles": 1086,
      "firstOffset": 86076,
      "lastOffset": 87162
    }
  ],
  "samples": []
}
