Files
rocm-systems/tests/workloads/ipblocks_SQ_TA/MI200/pmc_perf.csv
T
Jose Santos 35ef8c0707 update workloads]
Signed-off-by: Jose Santos <josantos@amd.com>
2024-03-07 15:26:40 -06:00

5.2 KiB

1Dispatch_IDKernel_NameGPU_IDGrid_SizeWorkgroup_SizeLDS_Per_WorkgroupScratch_Per_WorkitemArch_VGPRAccum_VGPRSGPRwave_sizeobjSQ_WAVES_LT_32SQ_WAVES_LT_16SQ_ITEMSSQ_LDS_MEM_VIOLATIONSSQ_LDS_ATOMIC_RETURNSQ_LDS_IDX_ACTIVESQ_WAVES_RESTOREDSQ_WAVES_SAVEDTA_FLAT_WRITE_WAVEFRONTS_sumTA_FLAT_ATOMIC_WAVEFRONTS_sumwave_size_1obj_1SQC_TC_DATA_WRITE_REQSQC_TC_DATA_ATOMIC_REQSQC_TC_STALLSQC_TC_REQSQC_DCACHE_REQ_READ_16SQC_ICACHE_REQSQC_ICACHE_HITSSQC_ICACHE_MISSESwave_size_2obj_2SQ_IFETCHSQ_LDS_BANK_CONFLICTSQ_LDS_ADDR_CONFLICTSQ_LDS_UNALIGNED_STALLSQ_WAVESSQ_WAVES_EQ_64SQ_WAVES_LT_64SQ_WAVES_LT_48TA_FLAT_WAVEFRONTS_sumTA_FLAT_READ_WAVEFRONTS_sumwave_size_3obj_3SQ_ACTIVE_INST_ANYSQ_CYCLESSQ_BUSY_CYCLESSQ_BUSY_CU_CYCLESSQ_ACTIVE_INST_VMEMSQ_ACTIVE_INST_LDSSQ_ACTIVE_INST_VALUSQ_ACTIVE_INST_SCATA_ADDR_STALLED_BY_TC_CYCLES_sumTA_TOTAL_WAVEFRONTS_sumwave_size_4obj_4SQ_INSTS_VALU_CVTSQ_INSTS_VMEM_WRSQ_INSTS_VMEM_RDSQ_INSTS_VMEMSQ_INSTS_SALUSQ_INSTS_VSKIPPEDSQ_INSTSSQ_INSTS_VALUGRBM_COUNTGRBM_GUI_ACTIVETA_TA_BUSY_sumTA_BUFFER_WAVEFRONTS_sumwave_size_5obj_5SQ_INSTS_LDSSQ_INSTS_GDSSQ_INSTS_EXP_GDSSQ_INSTS_BRANCHSQ_INSTS_SENDMSGSQ_WAVE_CYCLESSQ_WAIT_ANYSQ_WAIT_INST_ANYTA_BUFFER_COALESCED_READ_CYCLES_sumTA_BUFFER_COALESCED_WRITE_CYCLES_sumwave_size_6obj_6SQC_DCACHE_REQ_READ_1SQC_DCACHE_REQ_READ_2SQC_DCACHE_REQ_READ_4wave_size_7obj_7SQ_ACTIVE_INST_EXP_GDSSQ_ACTIVE_INST_MISCSQ_ACTIVE_INST_FLATSQ_INST_CYCLES_VMEM_WRSQ_INST_CYCLES_VMEM_RDSQ_INST_CYCLES_SMEMSQ_INST_CYCLES_SALUSQ_THREAD_CYCLES_VALUTA_ADDR_STALLED_BY_TD_CYCLES_sumTA_DATA_STALLED_BY_TC_CYCLES_sumwave_size_8obj_8SQ_INSTS_VALU_ADD_F64SQ_INSTS_VALU_MUL_F64SQ_INSTS_VALU_FMA_F64SQ_INSTS_VALU_TRANS_F64SQ_INSTS_VALU_INT32SQ_INSTS_VALU_INT64SQ_INSTS_SMEMSQ_INSTS_FLATTA_BUFFER_ATOMIC_WAVEFRONTS_sumTA_BUFFER_TOTAL_CYCLES_sumwave_size_9obj_9SQ_INSTS_VALU_ADD_F16SQ_INSTS_VALU_MUL_F16SQ_INSTS_VALU_FMA_F16SQ_INSTS_VALU_TRANS_F16SQ_INSTS_VALU_ADD_F32SQ_INSTS_VALU_MUL_F32SQ_INSTS_VALU_FMA_F32SQ_INSTS_VALU_TRANS_F32TA_BUFFER_READ_WAVEFRONTS_sumTA_BUFFER_WRITE_WAVEFRONTS_sumwave_size_10obj_10SQ_INSTS_FLAT_LDS_ONLYSQ_INSTS_VALU_MFMA_MOPS_I8SQ_INSTS_VALU_MFMA_MOPS_F16SQ_INSTS_VALU_MFMA_MOPS_BF16SQ_INSTS_VALU_MFMA_MOPS_F32SQ_INSTS_VALU_MFMA_MOPS_F64SQC_TC_INST_REQSQC_TC_DATA_READ_REQwave_size_11obj_11SQC_ICACHE_MISSES_DUPLICATESQC_DCACHE_INPUT_VALID_READYBSQC_DCACHE_ATOMICSQC_DCACHE_REQ_READ_8SQC_DCACHE_REQSQC_DCACHE_HITSSQC_DCACHE_MISSESSQC_DCACHE_MISSES_DUPLICATEwave_size_12obj_12SQ_INSTS_SMEM_NORMSQ_INSTS_MFMASQ_INSTS_VALU_MFMA_I8SQ_INSTS_VALU_MFMA_F16SQ_INSTS_VALU_MFMA_BF16SQ_INSTS_VALU_MFMA_F32SQ_INSTS_VALU_MFMA_F64SQ_VALU_MFMA_BUSY_CYCLESStart_TimestampEnd_Timestamp
20vecCopy(double*, double*, double*, int, int) 21048576256008016640x7f7151634ec00010485760000016384.00.0640x7f61b0404ec00002240655366420956640x7f84a35c4ec06553600016384163840032768.016384.0640x7fa133074ec0344064226464134756170262100163840114688259449.032768.0640x7ff1a5bd8ec0016384163843276849152039321616384027784277841038475.00.0640x7fb51c518ec000016384163841061285395128447559450.00.0640x7fdf34ae4ec032768327680640x7f4c20ba8ec00327683276816384163846553649152104857600.0225616.0640x7f0416f28ec00000114688065536327680.00.0640x7fdd9bf3cec0000000000.00.0640x7f0a3d084ec000000016856640x7fb675a04ec01308119573006553663800561680640x7f142814cec01310720000000197636385194490197636385214970
31vecCopy(double*, double*, double*, int, int) 21048576256008016640x7f7151634ec00010485760000016384.00.0640x7f61b0404ec000056065536655360640x7f84a35c4ec06553600016384163840032768.016384.0640x7fa133074ec0344064299936206150255898900163840114688446443.032768.0640x7ff1a5bd8ec0016384163843276849152039321616384038484384841736112.00.0640x7fb51c518ec0000163841638417686698171041222385120.00.0640x7fdf34ae4ec032768327680640x7f4c20ba8ec00327683276816384163846553649152104857600.0470988.0640x7f0416f28ec00000114688065536327680.00.0640x7fdd9bf3cec0000000000.00.0640x7f0a3d084ec0000000056640x7fb675a04ec00192083006553663800561680640x7f142814cec01310720000000197636385234650197636385250330
42vecCopy(double*, double*, double*, int, int) 21048576256008016640x7f7151634ec00010485760000016384.00.0640x7f61b0404ec000056065536655360640x7f84a35c4ec06553600016384163840032768.016384.0640x7fa133074ec0344064337144231351288637000163840114688625844.032768.0640x7ff1a5bd8ec0016384163843276849152039321616384040532405321638787.00.0640x7fb51c518ec00001638416384198616261850582810117340.00.0640x7fdf34ae4ec032768327680640x7f4c20ba8ec00327683276816384163846553649152104857600.0492901.0640x7f0416f28ec00000114688065536327680.00.0640x7fdd9bf3cec0000000000.00.0640x7f0a3d084ec0000000056640x7fb675a04ec00190677006553663800561680640x7f142814cec01310720000000197636385313530197636385329850