Task03 Евгений Свирин ITMO

[EvgenySvirin]

Локальный вывод

aplusb:
OpenCL devices:
  Device #0: GPU. NVIDIA GeForce RTX 4060 Laptop GPU. Total memory: 7940 Mb
Using device #0: GPU. NVIDIA GeForce RTX 4060 Laptop GPU. Total memory: 7940 Mb
Data generated for n=50000000!
Just example of printf usage: WARP_SIZE=32

mandelbrot:
OpenCL devices:
  Device #0: GPU. NVIDIA GeForce RTX 4060 Laptop GPU. Total memory: 7940 Mb
Using device #0: GPU. NVIDIA GeForce RTX 4060 Laptop GPU. Total memory: 7940 Mb
CPU: 0.286341+-0.00418889 s
CPU: 34.9234 GFlops
    Real iterations fraction: 56.2638%
GPU: 0.002554+-1.1547e-06 s
GPU: 4204.16 GFlops
    Real iterations fraction: 56.2657%
GPU vs CPU average results difference: 0.942446%

sum:
CPU:     0.117916+-0.0015682 s
CPU:     848.058 millions/s
CPU OMP: 0.0202012+-0.00288981 s
CPU OMP: 4950.21 millions/s
OpenCL devices:
  Device #0: GPU. NVIDIA GeForce RTX 4060 Laptop GPU. Total memory: 7940 Mb
Using device #0: GPU. NVIDIA GeForce RTX 4060 Laptop GPU. Total memory: 7940 Mb
GPU atomicSum1: 0.00332333+-1.10554e-06 s
GPU atomicSum1: 30090.3 millions/s
GPU loopSum2: 0.00980433+-0.000403407 s
GPU loopSum2: 10199.6 millions/s
GPU loopCoalescedSum3: 0.00934033+-9.42809e-07 s
GPU loopCoalescedSum3: 10706.3 millions/s
GPU localMemSum4: 0.00288017+-1.06719e-06 s
GPU localMemSum4: 34720.2 millions/s
GPU treeSum5: 0.00490683+-3.72678e-07 s
GPU treeSum5: 20379.7 millions/s

Вывод Github CI

aplusb:
Run ./build/aplusb
OpenCL devices:
  Device #0: CPU. AMD EPYC 7763 64-Core Processor                . Intel(R) Corporation. Total memory: 15991 Mb
Using device #0: CPU. AMD EPYC 7763 64-Core Processor                . Intel(R) Corporation. Total memory: 15991 Mb
Data generated for n=50000000!
Just example of printf usage: WARP_SIZE=1

mandelbrot:
Run ./mandelbrot
OpenCL devices:
  Device #0: CPU. AMD EPYC 7763 64-Core Processor                . Intel(R) Corporation. Total memory: 15991 Mb
Using device #0: CPU. AMD EPYC 7763 64-Core Processor                . Intel(R) Corporation. Total memory: 15991 Mb
CPU: 0.603726+-0.000772361 s
CPU: 16.5638 GFlops
    Real iterations fraction: 56.2638%
GPU: 0.162314+-0.00028869 s
GPU: 66.1522 GFlops
    Real iterations fraction: 56.2663%
GPU vs CPU average results difference: 0.982458%

sum:
Run ./sum
CPU:     0.0322993+-0.000169735 s
CPU:     3096.04 millions/s
CPU OMP: 0.0179648+-0.000388956 s
CPU OMP: 5566.43 millions/s
OpenCL devices:
  Device #0: CPU. AMD EPYC 7763 64-Core Processor                . Intel(R) Corporation. Total memory: 15991 Mb
Using device #0: CPU. AMD EPYC 7763 64-Core Processor                . Intel(R) Corporation. Total memory: 15991 Mb
GPU atomicSum1: 1.41342+-0.00113643 s
GPU atomicSum1: 70.7503 millions/s
GPU loopSum2: 1.69654+-0.00157669 s
GPU loopSum2: 58.9437 millions/s
GPU loopCoalescedSum3: 1.41982+-0.00104539 s
GPU loopCoalescedSum3: 70.4316 millions/s
GPU localMemSum4: 0.0284178+-3.72622e-05 s
GPU localMemSum4: 3518.92 millions/s
GPU treeSum5: 0.116099+-0.00186911 s
GPU treeSum5: 861.335 millions/s

[simiyutin]

В этой и следующей версии один поток выполняет больше работы, сказывается ли это как-то на конфигурации рабочего пространства? (а конфигурация на производительности)