[SYCL]rm wait() to improve the performance #7233
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
mofosyne
added
Review Complexity : Medium
|
can you paste the absolute performance number here? |
airMeng
approved these changes
May 13, 2024
teleprint-me
pushed a commit
to teleprint-me/llama.cpp
that referenced
this pull request
May 17, 2024
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Outdated suggestions cannot be applied.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.
This PR is used to revert the workaround solution in #5895.
That was a workaround to fix a known issue of oneMKL in Intel MTL Arc GPU.
Now, looks like the new oneMKL (oneAPI base toolkit 2024.1) is fixed the issue.
So, revert the old solution.
Now, we get the +32% in Intel MTL Arc GPU and +21% in Arc 770, tested with llama2-7b-Q4.
Next token:
MTL
7.06 tokens per second -> 9.37 tokens per second
Arc770
25.14 tokens per second ->30.50 tokens per second