{
  "clockHz": 1500000000,
  "tileCount": 1472,
  "sampleCount": 41294,
  "matchedSampleCount": 26574,
  "profileSpanCycles": 276990,
  "profileSpanMs": 0.18466,
  "atOffset": null,
  "groups": [
    {
      "name": "ipu_stack_gemm_f16_init_large_rows_interleaved_k384_c16_r114_r116",
      "dimensions": {},
      "sampleCount": 2484,
      "tileCount": 828,
      "phaseCount": 6,
      "phaseCycles": 55446,
      "phaseMs": 0.036964,
      "workCycles": 37856160,
      "averageActiveTiles": 682.7572773509361,
      "usefulCycles": 27398736.0,
      "estimatedWorkCycles": 37856160,
      "usefulUtilization": 0.7237589866484081,
      "workLaneOccupancy": 0.9904891304347826,
      "deviceUsefulUtilization": 0.3357008934662633,
      "meanCycles": 15240.0,
      "p50Cycles": 15240,
      "p95Cycles": 15240,
      "maximumCycles": 15240,
      "firstOffset": 300,
      "lastOffset": 73158
    },
    {
      "name": "ipu_stack_gemm_f16_init_small_rows_interleaved_k384_c16_r114_r116",
      "dimensions": {},
      "sampleCount": 1656,
      "tileCount": 552,
      "phaseCount": 4,
      "phaseCycles": 55236,
      "phaseMs": 0.036824,
      "workCycles": 25108272,
      "averageActiveTiles": 454.56354551379536,
      "usefulCycles": 17950896.0,
      "estimatedWorkCycles": 25108272,
      "usefulUtilization": 0.7149395227198431,
      "workLaneOccupancy": 0.9904891304347826,
      "deviceUsefulUtilization": 0.220778155078446,
      "meanCycles": 15162.0,
      "p50Cycles": 15162,
      "p95Cycles": 15162,
      "maximumCycles": 15162,
      "firstOffset": 306,
      "lastOffset": 73092
    },
    {
      "name": "ipu_stack_gemm_f16_init_large_rows_k80_c768_r7_r8",
      "dimensions": {},
      "sampleCount": 1360,
      "tileCount": 1360,
      "phaseCount": 6,
      "phaseCycles": 54852,
      "phaseMs": 0.036568,
      "workCycles": 74451840,
      "averageActiveTiles": 1357.3222489608402,
      "usefulCycles": 8922960.0,
      "estimatedWorkCycles": 74451840,
      "usefulUtilization": 0.11984875054800526,
      "workLaneOccupancy": 0.854296875,
      "deviceUsefulUtilization": 0.11051180409576442,
      "meanCycles": 54744.0,
      "p50Cycles": 54744,
      "p95Cycles": 54744,
      "maximumCycles": 54744,
      "firstOffset": 137754,
      "lastOffset": 192606
    },
    {
      "name": "ipu_stack_gemm_f16_init_small_rows_k80_c768_r7_r8",
      "dimensions": {},
      "sampleCount": 112,
      "tileCount": 112,
      "phaseCount": 2,
      "phaseCycles": 54390,
      "phaseMs": 0.03626,
      "workCycles": 6082944,
      "averageActiveTiles": 111.83938223938225,
      "usefulCycles": 642978.0,
      "estimatedWorkCycles": 6082944,
      "usefulUtilization": 0.10570177861246133,
      "workLaneOccupancy": 0.854296875,
      "deviceUsefulUtilization": 0.00803099294947121,
      "meanCycles": 54312.0,
      "p50Cycles": 54312,
      "p95Cycles": 54312,
      "maximumCycles": 54312,
      "firstOffset": 137772,
      "lastOffset": 192162
    },
    {
      "name": "ipu_stack_attention_softmax_d72_p768_tail_f16",
      "dimensions": {},
      "sampleCount": 1472,
      "tileCount": 1472,
      "phaseCount": 7,
      "phaseCycles": 40848,
      "phaseMs": 0.027232,
      "workCycles": 59359872,
      "averageActiveTiles": 1453.1891891891892,
      "usefulCycles": 34070544.0,
      "estimatedWorkCycles": 59359872,
      "usefulUtilization": 0.573965927689332,
      "workLaneOccupancy": 0.9986324786324786,
      "deviceUsefulUtilization": 0.566631169212691,
      "meanCycles": 40326.0,
      "p50Cycles": 40326,
      "p95Cycles": 40326,
      "maximumCycles": 40326,
      "firstOffset": 192084,
      "lastOffset": 232932
    },
    {
      "name": "ipu_stack_copy_u64",
      "dimensions": {},
      "sampleCount": 4957,
      "tileCount": 1472,
      "phaseCount": 32,
      "phaseCycles": 22668,
      "phaseMs": 0.015112,
      "workCycles": 4080594,
      "averageActiveTiles": 180.01561672842774,
      "usefulCycles": 0.0,
      "estimatedWorkCycles": 0,
      "usefulUtilization": null,
      "workLaneOccupancy": null,
      "deviceUsefulUtilization": null,
      "meanCycles": 823.1983054266693,
      "p50Cycles": 570,
      "p95Cycles": 1230,
      "maximumCycles": 2010,
      "firstOffset": 20970,
      "lastOffset": 274014
    },
    {
      "name": "ipu_stack_gemm_f16_init_small_rows_interleaved_k96_c80_r91_r92",
      "dimensions": {},
      "sampleCount": 896,
      "tileCount": 896,
      "phaseCount": 7,
      "phaseCycles": 19914,
      "phaseMs": 0.013276,
      "workCycles": 14477568,
      "averageActiveTiles": 727.0045194335644,
      "usefulCycles": 8358714.0,
      "estimatedWorkCycles": 14477568,
      "usefulUtilization": 0.5773562244708503,
      "workLaneOccupancy": 0.854296875,
      "deviceUsefulUtilization": 0.28514985360965195,
      "meanCycles": 16158.0,
      "p50Cycles": 16158,
      "p95Cycles": 16158,
      "maximumCycles": 16158,
      "firstOffset": 241680,
      "lastOffset": 261594
    },
    {
      "name": "ipu_stack_gemm_f16_init_large_rows_interleaved_k96_c80_r91_r92",
      "dimensions": {},
      "sampleCount": 128,
      "tileCount": 128,
      "phaseCount": 4,
      "phaseCycles": 18636,
      "phaseMs": 0.012424,
      "workCycles": 2072832,
      "averageActiveTiles": 111.2273019961365,
      "usefulCycles": 1207224.0,
      "estimatedWorkCycles": 2072832,
      "usefulUtilization": 0.5824032048907003,
      "workLaneOccupancy": 0.854296875,
      "deviceUsefulUtilization": 0.04400756600128783,
      "meanCycles": 16194.0,
      "p50Cycles": 16194,
      "p95Cycles": 16194,
      "maximumCycles": 16194,
      "firstOffset": 243138,
      "lastOffset": 261774
    },
    {
      "name": "ipu_stack_reduce_sum_f16",
      "dimensions": {},
      "sampleCount": 2852,
      "tileCount": 1472,
      "phaseCount": 18,
      "phaseCycles": 18450,
      "phaseMs": 0.0123,
      "workCycles": 8673840,
      "averageActiveTiles": 470.1268292682927,
      "usefulCycles": 2904768.0,
      "estimatedWorkCycles": 8673840,
      "usefulUtilization": 0.33488835394704075,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.10695652173913044,
      "meanCycles": 3041.3183730715286,
      "p50Cycles": 2568,
      "p95Cycles": 3792,
      "maximumCycles": 3792,
      "firstOffset": 21846,
      "lastOffset": 271704
    },
    {
      "name": "ipu_stack_unpack_amp_to_row_major_f16_o1_r9_p16_c116_p116",
      "dimensions": {},
      "sampleCount": 12,
      "tileCount": 12,
      "phaseCount": 3,
      "phaseCycles": 14550,
      "phaseMs": 0.0097,
      "workCycles": 147456,
      "averageActiveTiles": 10.13443298969072,
      "usefulCycles": 6264.0,
      "estimatedWorkCycles": 147456,
      "usefulUtilization": 0.04248046875,
      "workLaneOccupancy": 0.5625,
      "deviceUsefulUtilization": 0.0002924697445091887,
      "meanCycles": 12288.0,
      "p50Cycles": 12288,
      "p95Cycles": 12288,
      "maximumCycles": 12288,
      "firstOffset": 74382,
      "lastOffset": 88932
    },
    {
      "name": "ipu_stack_unpack_amp_to_row_major_f16_o1_r9_p16_c114_p114",
      "dimensions": {},
      "sampleCount": 8,
      "tileCount": 8,
      "phaseCount": 2,
      "phaseCycles": 14274,
      "phaseMs": 0.009516,
      "workCycles": 96672,
      "averageActiveTiles": 6.772593526691887,
      "usefulCycles": 4104.0,
      "estimatedWorkCycles": 96672,
      "usefulUtilization": 0.04245283018867924,
      "workLaneOccupancy": 0.5625,
      "deviceUsefulUtilization": 0.00019532320850923845,
      "meanCycles": 12084.0,
      "p50Cycles": 12084,
      "p95Cycles": 12084,
      "maximumCycles": 12084,
      "firstOffset": 74418,
      "lastOffset": 88692
    },
    {
      "name": "ipu_stack_fill_zero_u64",
      "dimensions": {},
      "sampleCount": 4240,
      "tileCount": 1472,
      "phaseCount": 24,
      "phaseCycles": 14004,
      "phaseMs": 0.009336,
      "workCycles": 1712928,
      "averageActiveTiles": 122.31705227077978,
      "usefulCycles": 0.0,
      "estimatedWorkCycles": 526848,
      "usefulUtilization": 0.0,
      "workLaneOccupancy": null,
      "deviceUsefulUtilization": 0.0,
      "meanCycles": 403.99245283018865,
      "p50Cycles": 378,
      "p95Cycles": 588,
      "maximumCycles": 3276,
      "firstOffset": 73434,
      "lastOffset": 273504
    },
    {
      "name": "ipu_stack_unpack_amp_to_row_major_f16_o1_r16_p16_c116_p116",
      "dimensions": {},
      "sampleCount": 540,
      "tileCount": 540,
      "phaseCount": 3,
      "phaseCycles": 13206,
      "phaseMs": 0.008804,
      "workCycles": 5883840,
      "averageActiveTiles": 445.54293502953203,
      "usefulCycles": 501120.0,
      "estimatedWorkCycles": 5883840,
      "usefulUtilization": 0.08516886930983847,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.025778796199356024,
      "meanCycles": 10896.0,
      "p50Cycles": 10896,
      "p95Cycles": 10896,
      "maximumCycles": 10896,
      "firstOffset": 74376,
      "lastOffset": 87582
    },
    {
      "name": "ipu_stack_unpack_amp_to_row_major_f16_o1_r16_p16_c114_p114",
      "dimensions": {},
      "sampleCount": 360,
      "tileCount": 360,
      "phaseCount": 2,
      "phaseCycles": 13002,
      "phaseMs": 0.008668,
      "workCycles": 3857760,
      "averageActiveTiles": 296.7051222888786,
      "usefulCycles": 328320.0,
      "estimatedWorkCycles": 3857760,
      "usefulUtilization": 0.0851063829787234,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.017154551473686323,
      "meanCycles": 10716.0,
      "p50Cycles": 10716,
      "p95Cycles": 10716,
      "maximumCycles": 10716,
      "firstOffset": 74382,
      "lastOffset": 87384
    },
    {
      "name": "ipu_stack_copy_strided_u64",
      "dimensions": {},
      "sampleCount": 1592,
      "tileCount": 1472,
      "phaseCount": 19,
      "phaseCycles": 12522,
      "phaseMs": 0.008348,
      "workCycles": 811668,
      "averageActiveTiles": 64.81935793004313,
      "usefulCycles": 0.0,
      "estimatedWorkCycles": 0,
      "usefulUtilization": null,
      "workLaneOccupancy": null,
      "deviceUsefulUtilization": null,
      "meanCycles": 509.84170854271355,
      "p50Cycles": 510,
      "p95Cycles": 510,
      "maximumCycles": 834,
      "firstOffset": 85668,
      "lastOffset": 274524
    },
    {
      "name": "ipu_stack_rearrange_row_major_to_amp_f16_o0_r8_p8_c80_p80",
      "dimensions": {},
      "sampleCount": 1360,
      "tileCount": 1360,
      "phaseCount": 5,
      "phaseCycles": 12282,
      "phaseMs": 0.008188,
      "workCycles": 3361920,
      "averageActiveTiles": 273.7274059599414,
      "usefulCycles": 435200.0,
      "estimatedWorkCycles": 3361920,
      "usefulUtilization": 0.12944983818770225,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.02407198940832466,
      "meanCycles": 2472.0,
      "p50Cycles": 2472,
      "p95Cycles": 2472,
      "maximumCycles": 2472,
      "firstOffset": 92538,
      "lastOffset": 104820
    },
    {
      "name": "ipu_stack_rearrange_row_major_to_amp_f16_o1_r64_p64_c16_p16",
      "dimensions": {},
      "sampleCount": 880,
      "tileCount": 880,
      "phaseCount": 4,
      "phaseCycles": 11310,
      "phaseMs": 0.00754,
      "workCycles": 1320000,
      "averageActiveTiles": 116.71087533156499,
      "usefulCycles": 450560.0,
      "estimatedWorkCycles": 1320000,
      "usefulUtilization": 0.3413333333333333,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.02706339138123246,
      "meanCycles": 1500.0,
      "p50Cycles": 1500,
      "p95Cycles": 1500,
      "maximumCycles": 1500,
      "firstOffset": 95010,
      "lastOffset": 106320
    },
    {
      "name": "ipu_stack_rearrange_row_major_to_amp_f16_o1_r0_p64_c0_p16",
      "dimensions": {},
      "sampleCount": 80,
      "tileCount": 80,
      "phaseCount": 2,
      "phaseCycles": 9180,
      "phaseMs": 0.00612,
      "workCycles": 216960,
      "averageActiveTiles": 23.633986928104576,
      "usefulCycles": 16000.0,
      "estimatedWorkCycles": 216960,
      "usefulUtilization": 0.07374631268436578,
      "workLaneOccupancy": 0.390625,
      "deviceUsefulUtilization": 0.0011840484986265038,
      "meanCycles": 2712.0,
      "p50Cycles": 2712,
      "p95Cycles": 2712,
      "maximumCycles": 2712,
      "firstOffset": 95310,
      "lastOffset": 104490
    },
    {
      "name": "ipu_stack_attention_merge_v72_p80_k768_f16",
      "dimensions": {},
      "sampleCount": 1472,
      "tileCount": 1472,
      "phaseCount": 14,
      "phaseCycles": 7614,
      "phaseMs": 0.005076,
      "workCycles": 3629952,
      "averageActiveTiles": 476.7470449172577,
      "usefulCycles": 0.0,
      "estimatedWorkCycles": 0,
      "usefulUtilization": null,
      "workLaneOccupancy": null,
      "deviceUsefulUtilization": null,
      "meanCycles": 2466.0,
      "p50Cycles": 2466,
      "p95Cycles": 2466,
      "maximumCycles": 2466,
      "firstOffset": 269376,
      "lastOffset": 276990
    },
    {
      "name": "ipu_stack_rearrange_row_major_to_amp_f16_o0_r7_p7_c80_p80",
      "dimensions": {},
      "sampleCount": 112,
      "tileCount": 112,
      "phaseCount": 2,
      "phaseCycles": 6444,
      "phaseMs": 0.004296,
      "workCycles": 276864,
      "averageActiveTiles": 42.96461824953445,
      "usefulCycles": 31360.0,
      "estimatedWorkCycles": 276864,
      "usefulUtilization": 0.11326860841423948,
      "workLaneOccupancy": 1.0,
      "deviceUsefulUtilization": 0.0033060750816398133,
      "meanCycles": 2472.0,
      "p50Cycles": 2472,
      "p95Cycles": 2472,
      "maximumCycles": 2472,
      "firstOffset": 91452,
      "lastOffset": 102792
    },
    {
      "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": 86406,
      "lastOffset": 87492
    }
  ],
  "samples": []
}
