Group nccl all reduce #9548
Merged
Group nccl all reduce #9548
Conversation
This file contains hidden or 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
panyx0718 force-pushed the group_nccl_all_reduce branch 2 times, most recently from 37257cf to 836f069 Compare panyx0718 requested review from chengduoZH and reyoung and removed request for reyoung panyx0718 force-pushed the group_nccl_all_reduce branch 2 times, most recently from 114a6b2 to 4a76d1d Compare 
panyx0718 force-pushed the group_nccl_all_reduce branch 4 times, most recently from be193ab to cb5d752 Compare panyx0718 force-pushed the group_nccl_all_reduce branch from cb5d752 to b25a9c4 Compare 
chengduoZH approved these changes Apr 3, 2018
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
2 participants
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 Outdated suggestions cannot be applied.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.
se-resnext on 4 device titan-x goes from ~1.25 to ~0.98
Some changes in the PR are due to pre-commit cpplint
The main issue is that nccl_all_reduce stream cannot overlap with computation streams.
some slow nccl_all_reduce block other gpus' streams