{
  "clockHz": 1500000000,
  "tileCount": 1472,
  "sampleCount": 31326,
  "matchedSampleCount": 18078,
  "profileSpanCycles": 320256,
  "profileSpanMs": 0.213504,
  "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": 73194
    },
    {
      "name": "ipu_stack_gemm_f16_init_small_rows_interleaved_k384_c16_r114_r116",
      "dimensions": {},
      "sampleCount": 1656,
      "tileCount": 552,
      "phaseCount": 4,
      "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": 73110
    },
    {
      "name": "ipu_stack_gemm_f16_init_large_rows_k80_c768_r7_r8",
      "dimensions": {},
      "sampleCount": 1360,
      "tileCount": 1360,
      "phaseCount": 6,
      "phaseCycles": 54828,
      "phaseMs": 0.036552,
      "workCycles": 74451840,
      "averageActiveTiles": 1357.9163930838258,
      "usefulCycles": 8922960.0,
      "estimatedWorkCycles": 74451840,
      "usefulUtilization": 0.11984875054800526,
      "workLaneOccupancy": 0.854296875,
      "deviceUsefulUtilization": 0.11056017870906963,
      "meanCycles": 54744.0,
      "p50Cycles": 54744,
      "p95Cycles": 54744,
      "maximumCycles": 54744,
      "firstOffset": 137778,
      "lastOffset": 192606
    },
    {
      "name": "ipu_stack_gemm_f16_init_small_rows_k80_c768_r7_r8",
      "dimensions": {},
      "sampleCount": 112,
      "tileCount": 112,
      "phaseCount": 2,
      "phaseCycles": 54354,
      "phaseMs": 0.036236,
      "workCycles": 6082944,
      "averageActiveTiles": 111.91345623137211,
      "usefulCycles": 642978.0,
      "estimatedWorkCycles": 6082944,
      "usefulUtilization": 0.10570177861246133,
      "workLaneOccupancy": 0.854296875,
      "deviceUsefulUtilization": 0.008036312074948286,
      "meanCycles": 54312.0,
      "p50Cycles": 54312,
      "p95Cycles": 54312,
      "maximumCycles": 54312,
      "firstOffset": 137802,
      "lastOffset": 192156
    },
    {
      "name": "ipu_stack_gemm_f16_init_large_rows_k768_c80_r7_r8",
      "dimensions": {},
      "sampleCount": 1360,
      "tileCount": 1360,
      "phaseCount": 6,
      "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": 263994,
      "lastOffset": 317790
    },
    {
      "name": "ipu_stack_gemm_f16_init_small_rows_k768_c80_r7_r8",
      "dimensions": {},
      "sampleCount": 112,
      "tileCount": 112,
      "phaseCount": 2,
      "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": 264006,
      "lastOffset": 317274
    },
    {
      "name": "ipu_stack_attention_softmax_d72_p768_tail_f16",
      "dimensions": {},
      "sampleCount": 1472,
      "tileCount": 1472,
      "phaseCount": 7,
      "phaseCycles": 40818,
      "phaseMs": 0.027212,
      "workCycles": 59359872,
      "averageActiveTiles": 1454.2572394531824,
      "usefulCycles": 34070544.0,
      "estimatedWorkCycles": 59359872,
      "usefulUtilization": 0.573965927689332,
      "workLaneOccupancy": 0.9986324786324786,
      "deviceUsefulUtilization": 0.567047626047332,
      "meanCycles": 40326.0,
      "p50Cycles": 40326,
      "p95Cycles": 40326,
      "maximumCycles": 40326,
      "firstOffset": 192114,
      "lastOffset": 232932
    },
    {
      "name": "ipu_stack_unpack_amp_to_row_major_f16_o1_r9_p16_c116_p116",
      "dimensions": {},
      "sampleCount": 12,
      "tileCount": 12,
      "phaseCount": 3,
      "phaseCycles": 14544,
      "phaseMs": 0.009696,
      "workCycles": 147456,
      "averageActiveTiles": 10.138613861386139,
      "usefulCycles": 6264.0,
      "estimatedWorkCycles": 147456,
      "usefulUtilization": 0.04248046875,
      "workLaneOccupancy": 0.5625,
      "deviceUsefulUtilization": 0.0002925904003443823,
      "meanCycles": 12288.0,
      "p50Cycles": 12288,
      "p95Cycles": 12288,
      "maximumCycles": 12288,
      "firstOffset": 74424,
      "lastOffset": 88968
    },
    {
      "name": "ipu_stack_unpack_amp_to_row_major_f16_o1_r9_p16_c114_p114",
      "dimensions": {},
      "sampleCount": 8,
      "tileCount": 8,
      "phaseCount": 2,
      "phaseCycles": 14316,
      "phaseMs": 0.009544,
      "workCycles": 96672,
      "averageActiveTiles": 6.752724224643755,
      "usefulCycles": 4104.0,
      "estimatedWorkCycles": 96672,
      "usefulUtilization": 0.04245283018867924,
      "workLaneOccupancy": 0.5625,
      "deviceUsefulUtilization": 0.00019475017311126498,
      "meanCycles": 12084.0,
      "p50Cycles": 12084,
      "p95Cycles": 12084,
      "maximumCycles": 12084,
      "firstOffset": 74442,
      "lastOffset": 88758
    },
    {
      "name": "ipu_stack_unpack_amp_to_row_major_f16_o1_r16_p16_c116_p116",
      "dimensions": {},
      "sampleCount": 540,
      "tileCount": 540,
      "phaseCount": 3,
      "phaseCycles": 13182,
      "phaseMs": 0.008788,
      "workCycles": 5883840,
      "averageActiveTiles": 446.35411925352753,
      "usefulCycles": 501120.0,
      "estimatedWorkCycles": 5883840,
      "usefulUtilization": 0.08516886930983847,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.02582573073954602,
      "meanCycles": 10896.0,
      "p50Cycles": 10896,
      "p95Cycles": 10896,
      "maximumCycles": 10896,
      "firstOffset": 74406,
      "lastOffset": 87588
    },
    {
      "name": "ipu_stack_unpack_amp_to_row_major_f16_o1_r16_p16_c114_p114",
      "dimensions": {},
      "sampleCount": 360,
      "tileCount": 360,
      "phaseCount": 2,
      "phaseCycles": 12984,
      "phaseMs": 0.008656,
      "workCycles": 3857760,
      "averageActiveTiles": 297.11645101663584,
      "usefulCycles": 328320.0,
      "estimatedWorkCycles": 3857760,
      "usefulUtilization": 0.0851063829787234,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.017178333199389214,
      "meanCycles": 10716.0,
      "p50Cycles": 10716,
      "p95Cycles": 10716,
      "maximumCycles": 10716,
      "firstOffset": 74430,
      "lastOffset": 87414
    },
    {
      "name": "ipu_stack_reduce_sum_f16",
      "dimensions": {},
      "sampleCount": 1380,
      "tileCount": 1380,
      "phaseCount": 6,
      "phaseCycles": 12318,
      "phaseMs": 0.008212,
      "workCycles": 5183280,
      "averageActiveTiles": 420.78908913784704,
      "usefulCycles": 1271808.0,
      "estimatedWorkCycles": 5183280,
      "usefulUtilization": 0.2453674121405751,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.07014125669751584,
      "meanCycles": 3756.0,
      "p50Cycles": 3792,
      "p95Cycles": 3792,
      "maximumCycles": 3792,
      "firstOffset": 21882,
      "lastOffset": 80292
    },
    {
      "name": "ipu_stack_rearrange_row_major_to_amp_f16_o0_r8_p8_c80_p80",
      "dimensions": {},
      "sampleCount": 1360,
      "tileCount": 1360,
      "phaseCount": 5,
      "phaseCycles": 12192,
      "phaseMs": 0.008128,
      "workCycles": 3361920,
      "averageActiveTiles": 275.748031496063,
      "usefulCycles": 435200.0,
      "estimatedWorkCycles": 3361920,
      "usefulUtilization": 0.12944983818770225,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.02424968618053178,
      "meanCycles": 2472.0,
      "p50Cycles": 2472,
      "p95Cycles": 2472,
      "maximumCycles": 2472,
      "firstOffset": 92616,
      "lastOffset": 104808
    },
    {
      "name": "ipu_stack_rearrange_row_major_to_amp_f16_o1_r64_p64_c16_p16",
      "dimensions": {},
      "sampleCount": 880,
      "tileCount": 880,
      "phaseCount": 4,
      "phaseCycles": 11220,
      "phaseMs": 0.00748,
      "workCycles": 1320000,
      "averageActiveTiles": 117.6470588235294,
      "usefulCycles": 450560.0,
      "estimatedWorkCycles": 1320000,
      "usefulUtilization": 0.3413333333333333,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.027280477408354646,
      "meanCycles": 1500.0,
      "p50Cycles": 1500,
      "p95Cycles": 1500,
      "maximumCycles": 1500,
      "firstOffset": 95088,
      "lastOffset": 106308
    },
    {
      "name": "ipu_stack_rearrange_row_major_to_amp_f16_o1_r0_p64_c0_p16",
      "dimensions": {},
      "sampleCount": 80,
      "tileCount": 80,
      "phaseCount": 2,
      "phaseCycles": 10290,
      "phaseMs": 0.00686,
      "workCycles": 216960,
      "averageActiveTiles": 21.08454810495627,
      "usefulCycles": 16000.0,
      "estimatedWorkCycles": 216960,
      "usefulUtilization": 0.07374631268436578,
      "workLaneOccupancy": 0.390625,
      "deviceUsefulUtilization": 0.0010563231503781638,
      "meanCycles": 2712.0,
      "p50Cycles": 2712,
      "p95Cycles": 2712,
      "maximumCycles": 2712,
      "firstOffset": 95262,
      "lastOffset": 105552
    },
    {
      "name": "ipu_stack_copy_u64",
      "dimensions": {},
      "sampleCount": 1565,
      "tileCount": 1380,
      "phaseCount": 10,
      "phaseCycles": 10062,
      "phaseMs": 0.006708,
      "workCycles": 1835922,
      "averageActiveTiles": 182.46094215861658,
      "usefulCycles": 0.0,
      "estimatedWorkCycles": 0,
      "usefulUtilization": null,
      "workLaneOccupancy": null,
      "deviceUsefulUtilization": null,
      "meanCycles": 1173.1130990415336,
      "p50Cycles": 1182,
      "p95Cycles": 1182,
      "maximumCycles": 1626,
      "firstOffset": 21006,
      "lastOffset": 87672
    },
    {
      "name": "ipu_stack_fill_zero_u64",
      "dimensions": {},
      "sampleCount": 1744,
      "tileCount": 1472,
      "phaseCount": 5,
      "phaseCycles": 7968,
      "phaseMs": 0.005312,
      "workCycles": 722208,
      "averageActiveTiles": 90.63855421686748,
      "usefulCycles": 0.0,
      "estimatedWorkCycles": 0,
      "usefulUtilization": null,
      "workLaneOccupancy": null,
      "deviceUsefulUtilization": null,
      "meanCycles": 414.11009174311926,
      "p50Cycles": 378,
      "p95Cycles": 492,
      "maximumCycles": 3276,
      "firstOffset": 73470,
      "lastOffset": 90414
    },
    {
      "name": "ipu_stack_rearrange_row_major_to_amp_f16_o0_r7_p7_c80_p80",
      "dimensions": {},
      "sampleCount": 112,
      "tileCount": 112,
      "phaseCount": 2,
      "phaseCycles": 6498,
      "phaseMs": 0.004332,
      "workCycles": 276864,
      "averageActiveTiles": 42.607571560480146,
      "usefulCycles": 31360.0,
      "estimatedWorkCycles": 276864,
      "usefulUtilization": 0.11326860841423948,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.0032786007734821416,
      "meanCycles": 2472.0,
      "p50Cycles": 2472,
      "p95Cycles": 2472,
      "maximumCycles": 2472,
      "firstOffset": 91236,
      "lastOffset": 104064
    },
    {
      "name": "ipu_stack_attention_merge_v72_p80_k768_f16",
      "dimensions": {},
      "sampleCount": 1472,
      "tileCount": 1472,
      "phaseCount": 7,
      "phaseCycles": 3012,
      "phaseMs": 0.002008,
      "workCycles": 3629952,
      "averageActiveTiles": 1205.1633466135459,
      "usefulCycles": 0.0,
      "estimatedWorkCycles": 0,
      "usefulUtilization": null,
      "workLaneOccupancy": null,
      "deviceUsefulUtilization": null,
      "meanCycles": 2466.0,
      "p50Cycles": 2466,
      "p95Cycles": 2466,
      "maximumCycles": 2466,
      "firstOffset": 317244,
      "lastOffset": 320256
    },
    {
      "name": "ipu_stack_copy_strided_u64",
      "dimensions": {},
      "sampleCount": 8,
      "tileCount": 8,
      "phaseCount": 2,
      "phaseCycles": 2118,
      "phaseMs": 0.001412,
      "workCycles": 6132,
      "averageActiveTiles": 2.895184135977337,
      "usefulCycles": 0.0,
      "estimatedWorkCycles": 0,
      "usefulUtilization": null,
      "workLaneOccupancy": null,
      "deviceUsefulUtilization": null,
      "meanCycles": 766.5,
      "p50Cycles": 834,
      "p95Cycles": 834,
      "maximumCycles": 834,
      "firstOffset": 85704,
      "lastOffset": 88410
    },
    {
      "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": 86454,
      "lastOffset": 87540
    }
  ],
  "samples": []
}
