Add thread yield function for spin-loops. #82
Closed
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
|
Is the 8% deviation reproducible within a single job, i.e. the same node over a short time interval? I'll see if I can reproduce locally. |
|
what happened to this? why wasn't merged? |
|
@gaming-hacker I haven't measured any performance improvement from |
|
FYI the OpenBLAS have been looking at the same issue, but don't have a clear conclusion: OpenMathLib/OpenBLAS#1051 |
4 tasks
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Outdated suggestions cannot be applied.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.
I noticed that TBLIS outperforms BLIS by about 30 GFLOP/s (400 vs. 370) on 12 cores of Lonestar 5 (both using OMP). To try and fix this I added a
pauseinstruction in the spin-loop for Intel architectures, but alas there is no improvement. In any case, the optional macroBLIS_YIELDthat I added inbli_kernel.hmight be useful.@fgvanzee, @tlrmchlsmth we should go over the respective codes at some point and figure out why TBLIS is faster.