Task05 Даниил Сухоруков ITMO

[UserNew35]

Локальный вывод

$ ./main_radix_sort 
Found 1 GPUs in 0.0320939 sec (OpenCL: 0.029603 sec, Vulkan: 0.00243806 sec)
Available devices:
  Device #0: API: OpenCL. CPU. AMD EPYC 7713P 64-Core Processor               . Intel(R) Corporation. Total memory: 9941 Mb.
Using device #0: API: OpenCL. CPU. AMD EPYC 7713P 64-Core Processor               . Intel(R) Corporation. Total memory: 9941 Mb.
Using OpenCL API...
n=100000000 max_value=2147483647
sorting on CPU...
CPU std::sort finished in 48.8895 sec
CPU std::sort effective RAM bandwidth: 0.0152396 GB/s (2.04542 uint millions/s)
Kernels compilation done in 0.148499 seconds
Kernels compilation done in 0.035999 seconds
Kernels compilation done in 0.0462752 seconds
Kernels compilation done in 0.0480339 seconds
GPU radix-sort times (in seconds) - 10 values (min=5.75503 10%=5.76577 median=5.82594 90%=6.2036 max=6.2036)
GPU radix-sort median effective VRAM bandwidth: 0.127886 GB/s (17.1646 uint millions/s)

Вывод Github CI

$ ./main_radix_sort 0
Found 2 GPUs in 0.0462139 sec (CUDA: 8.0511e-05 sec, OpenCL: 0.0219874 sec, Vulkan: 0.0240971 sec)
Available devices:
  Device #0: API: OpenCL. CPU. AMD EPYC 7763 64-Core Processor                . Intel(R) Corporation. Total memory: 15995 Mb.
  Device #1: API: Vulkan. CPU. llvmpipe (LLVM 20.1.2, 256 bits). Free memory: 15995/15995 Mb.
Using device #0: API: OpenCL. CPU. AMD EPYC 7763 64-Core Processor                . Intel(R) Corporation. Total memory: 15995 Mb.
Using OpenCL API...
n=100000000 max_value=2147483647
sorting on CPU...
CPU std::sort finished in 8.31966 sec
CPU std::sort effective RAM bandwidth: 0.0895537 GB/s (12.0197 uint millions/s)
Kernels compilation done in 0.148656 seconds
Kernels compilation done in 0.0414588 seconds
Kernels compilation done in 0.0424124 seconds
Kernels compilation done in 0.0444642 seconds
GPU radix-sort times (in seconds) - 10 values (min=4.6747 10%=4.68067 median=4.71507 90%=4.99414 max=4.99414)
GPU radix-sort median effective VRAM bandwidth: 0.158016 GB/s (21.2086 uint millions/s)