Poseidon: vectorized MDS matrix multiplication (AVX2) #229
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
nbgl
changed the title
dlubarov
reviewed
Sep 11, 2021
dlubarov
approved these changes
Sep 12, 2021
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Outdated suggestions cannot be applied.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.
AVX2 implementation of the MDS layer for Crandall. Makes width-8 Poseidon go from 1.8µs to 1.4µs and width-12 from 3.0µs to 2.1µs.
Yes, it’s completely unrolled. While I have heard of loops, they do not seem too useful here. Rust complains that
SHIFT_AMT[i]is not a constant (becauseiis a loop counter) and does not want to place it in an immediate. We could place some of the function in unrolled loops, but I actually found that harder to read and it was inexplicably slower.