[EPIC] CUDA Next asynchronous programming model
Open
cudax
- Dominant language
- C++
- Stars
- 2.5k
- Forks
- 486
- Avg merge
- 2d 6h
- Merged PRs (30d)
- 295
Description
What is the program representation of asynchronous work in CUDA Next?
### Tasks
- [ ] specify the interface for _safe_ asynchronous memory allocation
- [ ] specify an abstraction for an in-flight kernel and its products
- [ ] specify how asynchronous arguments get passed to kernel launches
- [ ] a way to specify data dependencies between kernels
- [ ] specify the representation of async memcpy's
Contributor guide
Assessment
This issue has not been assessed yet.