Files
rocm-systems/tests/workloads/vcopy/MI350/SQ_INST_LEVEL_VMEM.csv
T
vedithal-amd f9aa7be97c Support MI 350 profiling (#632)
* Add MI 350 hardware information

* Refactor MI GPU YAML file and corresponding interface

* Add SoC file for gfx950 architecture

* Add analysis report configs for MI 350 containing existing metrics

* Add placeholder None valued metrics for previous architectures to make
  baseline comparison work

* Enable testing on MI 350

* Analysis config metric changes
    - SPI changes
        - Update metric formula for default SPI pipe counter
             - Use efficiently collected pipe wise SPI counters
        - Add SPI Wave Occupancy
        - Add Scheduler-Pipe Wave Utilization
        - Update formula for VGPR Writes
        - Add Scheduler-Pipe FIFO Full Rate
   - CPC changes
	- Add CPC SYNC FIFO Full Rate
	- Add CPC CANE Stall Rate
        - Add CPC ADC Utilization
   - SQ changes
        - Add VALU co-issue efficiency
        - Add F6F4 datatype metrics
        - Update formula for total FLOPs by adding F6F4 counters
        - Add LDS STORE / LOAD / ATOMIC metrics
        - Add LDS STORE / LOAD / ATOMIC bandwidth
        - Add LDS FIFO and TA ADDR / CMD / DATA FIFO full rates

* Collect TCP_TCP_LATENCY_sum only for gfx950 (MI 350)

* Do not inject SQ_ACCUM_PREV_HIRES unnecesarily

* Do not hardcode memory and shader clock speeds

* Write num_hbm_channels to sysinfo.csv instead of hbm_bw while profiling

* Move generate sysinfo.csv to pre processing step of profiling

* Add warnings to use --specs-correction for missing sysinfo.csv values during analysis phase

* Update CHANGELOG

* Analysis phase warning to use --specs-correction when needed
2025-04-03 02:21:18 -04:00

9.5 KiB

1Dispatch_IDGPU_IDQueue_IDPIDTIDGrid_SizeWorkgroup_SizeLDS_Per_WorkgroupScratch_Per_WorkitemArch_VGPRAccum_VGPRSGPRWave_SizeKernel_NameStart_TimestampEnd_TimestampCorrelation_IDKernel_IDCPC_CANE_STALLCPC_CPC_STAT_STALLCPF_CPF_TCIU_BUSYCPF_CPF_TCIU_STALLGRBM_COUNTGRBM_SPI_BUSYSPI_CS0_CRAWLER_STALLSPI_CS1_CRAWLER_STALLSPI_CS2_CRAWLER_STALLSPI_CS3_CRAWLER_STALLSQ_ACTIVE_INST_VMEMSQ_INSTS_GDSSQ_INSTS_LDS_ATOMICSQ_INSTS_VALU_MFMA_F16SQ_INSTS_VALU_MFMA_MOPS_F32SQ_INSTS_VSKIPPEDSQ_INST_LEVEL_VMEMTA_ADDR_STALLED_BY_TD_CYCLES_sumTA_BUFFER_COALESCED_READ_CYCLES_sumTCC_EA0_WRREQ[0]TCC_EA0_WRREQ[10]TCC_EA0_WRREQ[11]TCC_EA0_WRREQ[12]TCC_EA0_WRREQ[14]TCC_EA0_WRREQ[15]TCC_EA0_WRREQ[16]TCC_EA0_WRREQ[17]TCC_EA0_WRREQ[18]TCC_EA0_WRREQ[19]TCC_EA0_WRREQ[1]TCC_EA0_WRREQ[21]TCC_EA0_WRREQ[22]TCC_EA0_WRREQ[23]TCC_EA0_WRREQ[24]TCC_EA0_WRREQ[25]TCC_EA0_WRREQ[26]TCC_EA0_WRREQ[27]TCC_EA0_WRREQ[28]TCC_EA0_WRREQ[2]TCC_EA0_WRREQ[30]TCC_EA0_WRREQ[31]TCC_EA0_WRREQ[32]TCC_EA0_WRREQ[33]TCC_EA0_WRREQ[34]TCC_EA0_WRREQ[35]TCC_EA0_WRREQ[36]TCC_EA0_WRREQ[37]TCC_EA0_WRREQ[38]TCC_EA0_WRREQ[39]TCC_EA0_WRREQ[3]TCC_EA0_WRREQ[41]TCC_EA0_WRREQ[42]TCC_EA0_WRREQ[43]TCC_EA0_WRREQ[44]TCC_EA0_WRREQ[45]TCC_EA0_WRREQ[46]TCC_EA0_WRREQ[47]TCC_EA0_WRREQ[48]TCC_EA0_WRREQ[49]TCC_EA0_WRREQ[50]TCC_EA0_WRREQ[51]TCC_EA0_WRREQ[52]TCC_EA0_WRREQ[53]TCC_EA0_WRREQ[54]TCC_EA0_WRREQ[55]TCC_EA0_WRREQ[57]TCC_EA0_WRREQ[58]TCC_EA0_WRREQ[59]TCC_EA0_WRREQ[5]TCC_EA0_WRREQ[60]TCC_EA0_WRREQ[61]TCC_EA0_WRREQ[62]TCC_EA0_WRREQ[63]TCC_EA0_WRREQ[6]TCC_EA0_WRREQ[7]TCC_EA0_WRREQ[8]TCC_EA0_WRREQ[9]TCC_MISS[0]TCC_MISS[10]TCC_MISS[11]TCC_MISS[12]TCC_MISS[14]TCC_MISS[15]TCC_MISS[16]TCC_MISS[17]TCC_MISS[18]TCC_MISS[19]TCC_MISS[1]TCC_MISS[21]TCC_MISS[22]TCC_MISS[23]TCC_MISS[24]TCC_MISS[25]TCC_MISS[26]TCC_MISS[27]TCC_MISS[28]TCC_MISS[2]TCC_MISS[30]TCC_MISS[31]TCC_MISS[32]TCC_MISS[33]TCC_MISS[34]TCC_MISS[35]TCC_MISS[36]TCC_MISS[37]TCC_MISS[38]TCC_MISS[39]TCC_MISS[3]TCC_MISS[41]TCC_MISS[42]TCC_MISS[43]TCC_MISS[44]TCC_MISS[45]TCC_MISS[46]TCC_MISS[47]TCC_MISS[48]TCC_MISS[49]TCC_MISS[50]TCC_MISS[51]TCC_MISS[52]TCC_MISS[53]TCC_MISS[54]TCC_MISS[55]TCC_MISS[57]TCC_MISS[58]TCC_MISS[59]TCC_MISS[5]TCC_MISS[60]TCC_MISS[61]TCC_MISS[62]TCC_MISS[63]TCC_MISS[6]TCC_MISS[7]TCC_MISS[8]TCC_MISS[9]TCC_READ[0]TCC_READ[10]TCC_READ[11]TCC_READ[12]TCC_READ[14]TCC_READ[15]TCC_READ[16]TCC_READ[17]TCC_READ[18]TCC_READ[19]TCC_READ[1]TCC_READ[21]TCC_READ[22]TCC_READ[23]TCC_READ[24]TCC_READ[25]TCC_READ[26]TCC_READ[27]TCC_READ[28]TCC_READ[2]TCC_READ[30]TCC_READ[31]TCC_READ[32]TCC_READ[33]TCC_READ[34]TCC_READ[35]TCC_READ[36]TCC_READ[37]TCC_READ[38]TCC_READ[39]TCC_READ[3]TCC_READ[41]TCC_READ[42]TCC_READ[43]TCC_READ[44]TCC_READ[45]TCC_READ[46]TCC_READ[47]TCC_READ[48]TCC_READ[49]TCC_READ[50]TCC_READ[51]TCC_READ[52]TCC_READ[53]TCC_READ[54]TCC_READ[55]TCC_READ[57]TCC_READ[58]TCC_READ[59]TCC_READ[5]TCC_READ[60]TCC_READ[61]TCC_READ[62]TCC_READ[63]TCC_READ[6]TCC_READ[7]TCC_READ[8]TCC_READ[9]TCC_WRITE[0]TCC_WRITE[10]TCC_WRITE[11]TCC_WRITE[12]TCC_WRITE[14]TCC_WRITE[15]TCC_WRITE[16]TCC_WRITE[17]TCC_WRITE[18]TCC_WRITE[19]TCC_WRITE[1]TCC_WRITE[21]TCC_WRITE[22]TCC_WRITE[23]TCC_WRITE[24]TCC_WRITE[25]TCC_WRITE[26]TCC_WRITE[27]TCC_WRITE[28]TCC_WRITE[2]TCC_WRITE[30]TCC_WRITE[31]TCC_WRITE[32]TCC_WRITE[33]TCC_WRITE[34]TCC_WRITE[35]TCC_WRITE[36]TCC_WRITE[37]TCC_WRITE[38]TCC_WRITE[39]TCC_WRITE[3]TCC_WRITE[41]TCC_WRITE[42]TCC_WRITE[43]TCC_WRITE[44]TCC_WRITE[45]TCC_WRITE[46]TCC_WRITE[47]TCC_WRITE[48]TCC_WRITE[49]TCC_WRITE[50]TCC_WRITE[51]TCC_WRITE[52]TCC_WRITE[53]TCC_WRITE[54]TCC_WRITE[55]TCC_WRITE[57]TCC_WRITE[58]TCC_WRITE[59]TCC_WRITE[5]TCC_WRITE[60]TCC_WRITE[61]TCC_WRITE[62]TCC_WRITE[63]TCC_WRITE[6]TCC_WRITE[7]TCC_WRITE[8]TCC_WRITE[9]TCP_READ_TAGCONFLICT_STALL_CYCLES_sumTCP_TCC_NC_WRITE_REQ_sumTCP_TCC_RW_READ_REQ_sumTCP_TCP_LATENCY_sumTD_COALESCABLE_WAVEFRONT_sumTD_SPI_STALL_sum
20011575815758104857625600401664vecCopy(double*, double*, double*, int, int)124822614081225612482261408337351130.016354.02108.00.074761.037009.00.00.00.00.00.00.00.00.00.00.0256862.00.00.04096.02048.02048.02048.02048.02048.04096.04096.02048.02048.04096.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.04096.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.04096.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.04098.02048.02048.02052.02050.02049.04097.04096.02050.02048.04098.02048.02048.02051.02048.02050.02048.02049.02048.02050.02052.02048.04102.02048.02050.02049.02048.02051.02050.02048.02048.02050.02049.02050.02048.02050.02048.02048.04096.02048.02051.02048.02049.02048.02050.02048.02052.02048.02048.02050.02048.02050.02048.02049.02048.02050.02049.02050.02050.01024.01024.01028.01096.01060.02066.02048.01096.01024.02050.01024.01024.01097.01024.01096.01024.01042.01024.01096.01098.01024.02054.01024.01096.01060.01024.01027.01096.01024.01024.01096.01060.01026.01024.01096.01024.01024.02048.01024.01097.01024.01042.01024.01096.01024.01098.01024.01024.01026.01024.01096.01024.01042.01024.01096.01060.01096.04096.02048.02048.02048.02048.02048.04096.04096.02048.02048.04096.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.04096.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.04096.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.00.00.065536.012901602.00.0601.0
31011575815758104857625600401664vecCopy(double*, double*, double*, int, int)124822614163250412482261416495032130.014445.02112.00.066766.029059.00.00.00.00.00.00.00.00.00.00.0186838.00.00.04096.02048.02048.02048.02048.02048.04096.04096.02048.02048.04096.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.04096.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.04096.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.04098.02048.02049.02048.02050.02049.04097.04096.02050.02048.04096.02049.02048.02050.02048.02050.02048.02048.02048.02050.02050.02048.04097.02048.02050.02049.02049.02048.02050.02048.02048.02050.02049.02049.02048.02050.02048.02049.04097.02048.02050.02048.02048.02048.02050.02048.02050.02048.02049.02049.02048.02050.02048.02048.02048.02050.02049.02050.02067.01024.01042.01024.01096.01060.02049.02048.01096.01024.02048.01025.01024.01096.01024.01096.01024.01024.01024.01096.01096.01024.02049.01024.01096.01060.01042.01024.01096.01024.01024.01096.01060.01025.01024.01096.01024.01042.02049.01024.01096.01024.01024.01024.01096.01024.01096.01024.01025.01025.01024.01096.01024.01024.01024.01096.01060.01096.04096.02048.02048.02048.02048.02048.04096.04096.02048.02048.04096.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.04096.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.04096.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.00.00.065536.08973198.00.0897.0
42011575815758104857625600401664vecCopy(double*, double*, double*, int, int)124822614235463712482261423717163130.011292.02055.00.063585.029133.00.00.00.00.00.00.00.00.00.00.0186316.00.00.04096.02048.02048.02048.02048.02048.04096.04096.02048.02048.04096.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.04096.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.04096.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.04097.02048.02048.02048.02050.02049.04096.04096.02050.02048.04096.02048.02048.02050.02048.02050.02048.02048.02048.02050.02050.02048.04097.02048.02050.02049.02048.02048.02050.02048.02048.02050.02049.02049.02048.02050.02048.02048.04096.02048.02050.02048.02048.02048.02050.02048.02050.02048.02048.02049.02048.02050.02048.02048.02048.02050.02049.02050.02066.01024.01024.01024.01096.01060.02048.02048.01096.01024.02048.01024.01024.01096.01024.01096.01024.01024.01024.01096.01096.01024.02066.01024.01096.01060.01024.01024.01096.01024.01024.01096.01060.01042.01024.01096.01024.01024.02048.01024.01096.01024.01024.01024.01096.01024.01096.01024.01024.01042.01024.01096.01024.01024.01024.01096.01060.01096.04096.02048.02048.02048.02048.02048.04096.04096.02048.02048.04096.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.04096.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.04096.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.02048.00.00.065536.08951646.00.0876.0