[WIP] experiment devirgo-style sumcheck + FRI fold #653
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
Collaborator
Author
|
Hi @chaosma @yczhangsjtu FYI, please refer changes on file "mpcs/src/basefold/commit_phase.rs", other are just noise and can be ignored. For benchmarn, before/after this change, e2e remain the same 😂
But the good thing is we can share sumcheck implementation between main proof/mpcs, so we can reduce technical burden and avoid duplicate implementation Hope this PR can help on WHIR implementation, if you need any input or review on WHIR, I will be happy to learn that and review your implementation |
3 tasks
Collaborator
Author
|
Closed in favour of #891 |
github-merge-queue bot
pushed a commit
that referenced
this pull request
Apr 3, 2025
Follow up on #843 Clean up basefold sumcheck and unified devirgo sumcheck with ceno. Previously in #653 before/after change is insignificant, but after bunch of refactor and optimised other critical path, now the benchmark shows promising result. ### Change highlights - [x] switch to ceno sumcheck and break down basefold sumcheck into 2 phases > we can't directly use ceno sumcheck as blackbox because of FRI part in basefold. - [x] and use evaluation form as uni-variates - [x] clean up old sumcheck impl ### benchmark With e2e Fibonacci 2^20 ``` fibonacci_max_steps_1048576/prove_fibonacci/fibonacci_max_steps_1048576 time: [3.1789 s 3.2022 s 3.2274 s] change: [-14.457% -13.632% -12.780%] (p = 0.00 < 0.05) Performance has improved. ``` Fibonacci 2^21 ``` fibonacci_max_steps_2097152/prove_fibonacci/fibonacci_max_steps_2097152 time: [5.2820 s 5.3010 s 5.3207 s] change: [-16.128% -15.753% -15.306%] (p = 0.00 < 0.05) Performance has improved. ``` Fibonacci 2^22 ``` fibonacci_max_steps_4194304/prove_fibonacci/fibonacci_max_steps_4194304 time: [10.369 s 10.397 s 10.423 s] change: [-10.485% -10.041% -9.5790%] (p = 0.00 < 0.05) Performance has improved. ```
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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.
An experiment to see how much gain we have on e2e latency
Also related to #568