Investigate potential `defer_temp_storage_destruction` optimization
Open
Nobody has claimed this yet.
cuda
enhancement
P2
- Dominant language
- C++
- Stars
- 2.4k
- Forks
- 270
- Avg merge
- 3d 6h
- Merged PRs (30d)
- 39
Description
Memory deallocation in defer_temp_storage_destruction might be blocking. We should investigate if launching all the destruction kernels in sequence (or fusing destructors in a single kernel) before releasing the memory is any better to address:
// TODO Bury all the memory associated with the stream provider and then
// deallocate the memory
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by locating defer_temp_storage_destruction and the TODO about burying stream-provider memory before deallocation. Investigate whether sequential destruction kernels or a fused destructor kernel improves behavior, and define completion through measured evidence before changing the implementation.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- cpp
- Domain
- performance
- Issue type
- Refactor
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100