Choose a more efficient algorithm. #16357
Merged
Conversation
This file contains 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
masterleinad
approved these changes
Dec 15, 2023
kronbichler
reviewed
Dec 16, 2023
Comment on lines
172
to
173
| // a factor of 2 if we broadcast it from the last process, which | ||
| // can compute it from its own prefix sum plus its own value. |
There was a problem hiding this comment.
It is usually not as bad as a factor of 2 at least for small sizes, because Allreduce typically only costs one single communication step more than the log(n) communications of Reduce or Bcast when seen from the most expensive rank. The algorithm is called recursive doubling and it makes all processes participate during the reduction, i.e., it does not try to minimize the number of communications over the whole machine, but the length of the dependency chain.
The change is still correct, but I would simply write that "it is cheaper if we broadcast from the last process".
There was a problem hiding this comment.
Fair enough. I've just removed the "by about a factor of 2" :-)
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
Follow-up to #16352. We can avoid an
MPI_Reducein favor of anMPI_Broadcast.