H2yaml hip

[TApplencourt]

No description provided.

[TApplencourt]

Checked with collen and hiplz

THAPI: Trace location: /home/applenco/thapi-traces/thapi_aggreg--2026-04-02T20:04:20+00:00
BACKEND_HIP,BACKEND_ZE | 1 Hostnames | 1 Processes | 1 Threads |

                           Name |     Time | Time(%) | Calls |  Average |      Min |      Max |
       __hipUnregisterFatBinary | 100.53ms |  48.46% |     1 | 100.53ms | 100.53ms | 100.53ms |
         __hipRegisterFatBinary |  99.44ms |  47.94% |     1 |  99.44ms |  99.44ms |  99.44ms |
                hipLaunchKernel |   2.80ms |   1.35% |     1 |   2.80ms |   2.80ms |   2.80ms |
                      hipMemcpy |   1.64ms |   0.79% |     3 | 547.44us | 254.69us |   1.05ms |
  zeCommandListAppendMemoryCopy |   1.08ms |   0.52% |     4 | 269.04us |  21.76us | 734.88us |
         zeEventHostSynchronize | 428.63us |   0.21% |     7 |  61.23us |    276ns | 150.52us |
           zeCommandQueueCreate | 251.22us |   0.12% |     1 | 251.22us | 251.22us | 251.22us |
                 zeModuleCreate | 241.10us |   0.12% |     1 | 241.10us | 241.10us | 241.10us |
                      zeMemFree | 155.59us |   0.08% |     5 |  31.12us |  19.55us |  44.30us |
zeCommandListAppendLaunchKernel | 115.94us |   0.06% |     6 |  19.32us |   7.38us |  75.47us |
                      hipMalloc | 100.29us |   0.05% |     2 |  50.15us |  27.42us |  72.88us |
    zeContextMakeMemoryResident |  89.52us |   0.04% |     5 |  17.90us |   9.63us |  36.66us |
               zeMemAllocShared |  85.85us |   0.04% |     1 |  85.85us |  85.85us |  85.85us |
              zeEventPoolCreate |  83.61us |   0.04% |     1 |  83.61us |  83.61us |  83.61us |
             zeEventPoolDestroy |  80.18us |   0.04% |     1 |  80.18us |  80.18us |  80.18us |
                        hipFree |  69.24us |   0.03% |     2 |  34.62us |  23.14us |  46.10us |
   zeCommandListCreateImmediate |  52.84us |   0.03% |     1 |  52.84us |  52.84us |  52.84us |
               zeMemAllocDevice |  47.52us |   0.02% |     5 |   9.50us |   4.36us |  16.54us |
                zeModuleDestroy |  40.27us |   0.02% |     1 |  40.27us |  40.27us |  40.27us |
                 zeEventDestroy |  18.93us |   0.01% |     6 |   3.15us |    686ns |  14.60us |
               zeEventHostReset |  13.52us |   0.01% |    10 |   1.35us |    224ns |   3.64us |
                 zeKernelCreate |  11.38us |   0.01% |     5 |   2.28us |   1.09us |   5.99us |
                  zeEventCreate |  11.23us |   0.01% |     6 |   1.87us |    372ns |   6.55us |
           zeCommandListDestroy |   9.33us |   0.00% |     1 |   9.33us |   9.33us |   9.33us |
                    zeDeviceGet |   9.21us |   0.00% |     2 |   4.61us |   1.86us |   7.35us |
     __hipPushCallConfiguration |   6.82us |   0.00% |     1 |   6.82us |   6.82us |   6.82us |
      zeCommandQueueSynchronize |   5.68us |   0.00% |    10 | 568.10ns |    211ns |   1.96us |
          __hipRegisterFunction |   4.42us |   0.00% |     1 |   4.42us |   4.42us |   4.42us |
                zeKernelDestroy |   3.75us |   0.00% |     5 | 750.60ns |    300ns |   2.09us |
              zeContextCreateEx |   3.17us |   0.00% |     1 |   3.17us |   3.17us |   3.17us |
               __hipRegisterVar |   3.17us |   0.00% |     2 |   1.58us |    601ns |   2.56us |
      zeKernelSetIndirectAccess |   2.87us |   0.00% |     6 | 477.50ns |    135ns |   1.58us |
           zeKernelSetGroupSize |   1.94us |   0.00% |     6 | 323.17ns |    159ns |    710ns |
               zeContextDestroy |   1.41us |   0.00% |     1 |   1.41us |   1.41us |   1.41us |
         zeModuleGetKernelNames |   1.31us |   0.00% |     2 | 653.00ns |    260ns |   1.05us |
      zeModuleBuildLogGetString |    965ns |   0.00% |     2 | 482.50ns |    317ns |    648ns |
          zeCommandQueueDestroy |    919ns |   0.00% |     1 | 919.00ns |    919ns |    919ns |
                    zeDriverGet |    855ns |   0.00% |     2 | 427.50ns |    261ns |    594ns |
                         zeInit |    802ns |   0.00% |     1 | 802.00ns |    802ns |    802ns |
      __hipPopCallConfiguration |    514ns |   0.00% |     1 | 514.00ns |    514ns |    514ns |
                          Total | 207.44ms | 100.00% |   121 |

Device profiling | 1 Hostnames | 1 Processes | 1 Threads | 1 Devices | 1 Subdevices |

                                 Name |     Time | Time(%) | Calls |  Average |      Min |      Max |
   zeCommandListAppendMemoryCopy(M2D) | 256.48us |  66.89% |     2 | 128.24us | 111.20us | 145.28us |
   zeCommandListAppendMemoryCopy(D2M) |  92.56us |  24.14% |     2 |  46.28us |   5.68us |  86.88us |
__chip_var_bind___chipspv_device_heap |   9.12us |   2.38% |     2 |   4.56us |   4.32us |   4.80us |
                                saxpy |   8.48us |   2.21% |     1 |   8.48us |   8.48us |   8.48us |
__chip_var_init___chipspv_device_heap |   6.72us |   1.75% |     1 |   6.72us |   6.72us |   6.72us |
             __chip_reset_non_symbols |   5.44us |   1.42% |     1 |   5.44us |   5.44us |   5.44us |
__chip_var_info___chipspv_device_heap |   4.64us |   1.21% |     1 |   4.64us |   4.64us |   4.64us |
                                Total | 383.44us | 100.00% |    10 |

Explicit memory traffic (BACKEND_ZE) | 1 Hostnames | 1 Processes | 1 Threads |

                              Name |    Byte | Byte(%) | Calls | Average |    Min |    Max |
       zeContextMakeMemoryResident |  8.39MB |  40.00% |     5 |  1.68MB |     8B | 4.19MB |
zeCommandListAppendMemoryCopy(M2D) |  8.39MB |  40.00% |     2 |  4.19MB | 4.19MB | 4.19MB |
zeCommandListAppendMemoryCopy(D2M) |  4.19MB |  20.00% |     2 |  2.10MB |    24B | 4.19MB |
                             Total | 20.97MB | 100.00% |     9 |
                             ```
                             
                             Look good