Add SVML AVX512 FP16 content#2
Merged
Merged
Conversation
Member
Author
|
@seberg Any objection to adding these new content? |
seberg
approved these changes
Mar 9, 2023
seberg
left a comment
Member
There was a problem hiding this comment.
Looks good to add (please go ahead if it helps with the other PR), seems "small" compared to the rest anyway. The NumPy PR should be the one where to discuss.
I am slightly hesitant about the order of things, but probably it doesn't matter. I.e. it would be nice to have have basic support for faster half before adding AVX512.
Member
|
Ah, nvm. I didn't realize we already did other speedups here anyway. |
Member
Author
|
Sounds good, thanks. |
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Open sourcing new content for FP16 umath functions based on AVX-512 FP16 ISA on Intel Sapphire Rapids. These were measured to be 140x faster than the scalar counterpart in NumPy (which required conversion to FP32 and back). These have max ULP error of 3.05 (detailed ULP error listed here: numpy/numpy#23351 (comment)).