deepseek-ai / deepseek-ai/DeepEP
Why is __syncwarp() needed in internode_ll kernel ?
Open
- Dominant language
- Cuda
- Stars
- 10.1k
- Forks
- 1.4k
- Avg merge
- 4d 1h
- Merged PRs (30d)
- 2
Description
Before the subsequent return or this_grid().sync(), why perform a syncwarp operation first?
Contributor guide
No contributing guide indexed for this repository
Assessment
This issue has not been assessed yet.