Autotuning kernel launch parameters
Nobody has claimed this yet.
- Dominant language
- Julia
- Stars
- 86
- Forks
- 12
- Avg merge
- 8h 24m
- Merged PRs (30d)
- 7
Description
There are often significant performance gains from e.g. increasing the tile size when a smaller type is used, or sometimes simply setting the entry hint occupancy=2 will have a big impact. It would be nice to be able to automatically find a good combination of parameters specific to the current array sizes, hardware, and arch.
See _autotuner.py with example usage in AttentionFMHA.py
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 reading experimental/src/cuda/tile_experimental/_autotuner.py and its example usage in samples/AttentionFMHA.py. Trace how launch parameters such as tile size and the occupancy entry hint are selected, then define how candidate combinations should be evaluated for array sizes, hardware, and architecture. Done means the system can automatically find and use a good parameter combination for the current workload.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- julia
- Domain
- hpc, performance
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100