Use caching allocators in Thrust #835
Merged
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
stephenswat
added
cuda
niermann999
approved these changes
Feb 3, 2025
I was very pleasantly surprised to learn that using our vecmem memory resources in Thrust is trivial! You simply create a polymorphic allocator from them and Thrust will accept them as-is. This should reduce the amount of memory allocations significantly throughput traccc. With this change, I observe a 10-15% improvement of the full-application throughput of traccc.
stephenswat
force-pushed
the
feat/thrust_cache
branch
from
bc4d5e5 to
13cd6af
Compare
|
krasznaa
reviewed
Feb 3, 2025
| measurements_view.ptr() + n_measurements, | ||
| measurement_sort_comp()); | ||
| thrust::sort( | ||
| thrust::cuda::par(std::pmr::polymorphic_allocator(&(m_mr.main))) |
There was a problem hiding this comment.
Strange that you needed to be so very explicit. 🤔 std::pmr::polymoprhic_allocator doesn't put explicit on its constructor for sure. That's how we can create vecmem::vector objects without always writing out the full allocator name in their constructors.
I guess I'll try thrust::cuda::par(&(m_mr.main)).on(stream) tomorrow, to see if that would also work...
|
Yeah - What a free lunch.. |
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Outdated suggestions cannot be applied.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.
I was very pleasantly surprised to learn that using our vecmem memory resources in Thrust is trivial! You simply create a polymorphic allocator from them and Thrust will accept them as-is. This should reduce the amount of memory allocations significantly throughput traccc.
With this change, I observe a 10-15% improvement of the full-application throughput of traccc. 🥳