Task03 Лопатин Алексей HSE #117
Open
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
simiyutin
reviewed
Oct 5, 2024
|
|
||
| { | ||
| timer t; | ||
| for (int iter = 0; iter < benchmarkingIters; ++iter) { |
There was a problem hiding this comment.
Здесь и далее измерение производительности GPU версий имеет не очень много смысла, потому что замеры включают в себя мажорирующее время на перекидывание массивов по шине
There was a problem hiding this comment.
(а еще и компиляцию кернела)
simiyutin
reviewed
Oct 5, 2024
| @@ -6,6 +6,8 @@ | |||
| #ifdef __unix__ | |||
| // X11-based | |||
| #define cimg_display 1 | |||
| # elif __APPLE__ | |||
There was a problem hiding this comment.
В итоге отображение у вас заработало?
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
mandelbrot
Локальный вывод
$ ./mandelbrot OpenCL devices: Device #0: GPU. Apple M1 Pro. Total memory: 10922 Mb Using device #0: GPU. Apple M1 Pro. Total memory: 10922 Mb CPU: 4.0721+-0.00862937 s CPU: 2.45573 GFlops Real iterations fraction: 56.2657% GPU: 0.0074115+-0.000379725 s GPU: 1349.25 GFlops Real iterations fraction: 56.2657% GPU vs CPU average results difference: 0%Вывод Github CI
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: 0.604094+-0.00315337 s CPU: 16.5537 GFlops Real iterations fraction: 56.2638% GPU: 0.162827+-0.000825261 s GPU: 61.4149 GFlops Real iterations fraction: 56.2657% GPU vs CPU average results difference: 0.943129%sum
Локальный вывод
Вывод Github CI
К сожалению, не удалось завести OpenMP на macos и увидеть легкий прирост производительности и программа работала на одном ядре. Локально выиграл вариант с деревом, однако, при увеличении
VALUES_PER_WORKITEMдо 128 выигрывает cycle + coalesced вариант, что связано с размером кэш линии на моем процессоре. Что интересно, даже atomic вариант локально приносит до 3х-кратного прироста производительности, вероятно из-за оптимизаций вроде локального кэша, как говорилось на лекции