Intel^® VTune^™ Profiler 2022.2.0

Recommendations:

EU Array Stalled/Idle: 71.0% of Elapsed time with GPU busy

GPU metrics detect some kernel issues. Use GPU Compute/Media Hotspots (preview) to understand how well your application runs on the specified hardware.

Execution % of Total Time: 46.7%

Execution time on the device is less than memory transfer time. Make sure your offload schema is optimal. Use Intel Advisor tool to get an insight into possible causes for inefficient offload.

Elapsed Time: 128.028s

GPU Time, % of Elapsed time: 85.4%

Top Hotspots when GPU was idle:

Function	Module	CPU Time
[Outside any known module]	[Unknown]	4.567s
[Skipped stack frame(s)]	[Unknown]	1.496s
Intel::OpenCL::Utils::AtomicCounter::operator long	libcpu_device_emu.so.2022.13.3.0	1.264s
std::__malloc_alloc::allocate	libstlport-dynamic.so	0.648s
_PyEval_EvalFrameDefault	python3.9	0.643s
[Others]	N/A	9.999s

Hottest Host Tasks:

Host Task	Task Time	% of Elapsed Time(%)	Task Count
clWaitForEvents	59.706s	46.6%	85,345
clEnqueueMemcpyINTEL	23.575s	18.4%	73,124
clEnqueueNDRangeKernel	4.470s	3.5%	12,221
tbb_parallel_for	3.832s	3.0%	42
clCreateContext	3.221s	2.5%	3
[Others]	3.243s	2.5%	48,905

Hottest GPU Computing Tasks:

Computing Task	Total Time	Execution Time	% of Total Time(%)	SIMD Width	Peak EU Threads Occupancy(%)	EU Threads Occupancy(%)	SIMD Utilization(%)
dppyPy_dppy_py_devfn__5F__5F_main_5F__5F__2E_calCentroidsSum2_24_4_2E_array_28_float32_2C__20_2d_2C__20_C_29__2E_array_28_int32_2C__20_1d_2C__20_C_29__2E_array_28_float32_2C__20_2d_2C__20_C_29__2E_array_28_int32_2C__20_1d_2C__20_C_29_	35.690s	22.697s	63.6%	8	100.0%	72.1%	100.0%
dppyPy_dppy_py_devfn__5F__5F_main_5F__5F__2E_groupByCluster_24_2_2E_array_28_float32_2C__20_2d_2C__20_C_29__2E_array_28_int32_2C__20_1d_2C__20_C_29__2E_array_28_float32_2C__20_2d_2C__20_C_29__2E_int64_2E_int64	26.339s	15.073s	57.2%	8	100.0%	90.2%	100.0%
[Outside any task]	9.684s	0s	0.0%
dppyPy_dppy_py_devfn__5F__5F_main_5F__5F__2E_updateCentroids_24_5_2E_array_28_float32_2C__20_2d_2C__20_C_29__2E_array_28_float32_2C__20_2d_2C__20_C_29__2E_array_28_int32_2C__20_1d_2C__20_C_29__2E_int64	5.421s	0.052s	1.0%	8	1.2%	0.1%	100.0%
dppyPy_dppy_py_devfn__5F__5F_main_5F__5F__2E_calCentroidsSum1_24_3_2E_array_28_float32_2C__20_2d_2C__20_C_29__2E_array_28_int32_2C__20_1d_2C__20_C_29_	3.746s	0.025s	0.7%	8	1.2%	0.0%	100.0%
[Others]	0.217s	0.001s	0.5%	N/A	N/A	N/A	N/A

Collection and Platform Info:
- Application Command Line: python "kmeans_kernel.py" "--steps" "1" "--size" "1048576" "--repeat" "100"
- Operating System: 5.4.0-80-generic DISTRIB_ID=Ubuntu DISTRIB_RELEASE=20.04 DISTRIB_CODENAME=focal DISTRIB_DESCRIPTION="Ubuntu 20.04.4 LTS"
- Computer Name: s001-n157
- Result Size: 6.5 GB
- Collection start time: 15:58:41 04/06/2022 UTC
- Collection stop time: 16:00:49 04/06/2022 UTC
- Collector Type: Event-based sampling driver,Driverless Perf system-wide sampling,User-mode sampling and tracing
- CPU:
  - Name: Intel(R) microarchitecture code named Coffeelake
  - Frequency: 3.696 GHz
  - Logical CPU Count: 12
- GPU:
  - Name: HD Graphics P630
  - Vendor: Intel Corporation
  - EU Count: 24
  - Max EU Thread Count: 7
  - Max Core Frequency: 1.200 GHz
  - GPU OpenCL Info:
    Version: OpenCL 3.0 NEO
    Max Compute Units: 24
    Max Work Group Size: 256
    Local Memory: 65.5 KB
    SVM Capabilities: