fix: optimize opening proof gpu usage #569
Open
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
sagar-a16z
force-pushed
the
sagar/optimized_opening
branch
3 times, most recently
from
d7344b6 to
6373f3f
Compare
sagar-a16z
changed the title
sagar-a16z
force-pushed
the
sagar/optimized_opening
branch
4 times, most recently
from
1e19b77 to
cc0b559
Compare
sagar-a16z
force-pushed
the
sagar/optimized_opening
branch
from
180d219 to
293fd54
Compare
moodlezoup
approved these changes
Jan 29, 2025
Comment on lines
+293
to
+295
| // let q_k_com: Vec<P::G1Affine> = optimal_iter!(quotients) | ||
| // .map(|q| UnivariateKZG::commit(&pp.commit_pp, q).unwrap()) | ||
| // .collect(); |
There was a problem hiding this comment.
nit
Suggested change
| // let q_k_com: Vec<P::G1Affine> = optimal_iter!(quotients) | |
| // .map(|q| UnivariateKZG::commit(&pp.commit_pp, q).unwrap()) | |
| // .collect(); |
| @@ -230,30 +230,66 @@ where | |||
| Ok(commitments.into_iter().map(|c| c.into_affine()).collect()) | |||
| } | |||
|
|
|||
| // This API will try to minimize copies to the GPU or just do the batches in parallel on the CPU | |||
| #[tracing::instrument(skip_all, name = "KZG::commit_variable_batch_with_mode")] | |||
There was a problem hiding this comment.
Suggested change
| #[tracing::instrument(skip_all, name = "KZG::commit_variable_batch_with_mode")] | |
| #[tracing::instrument(skip_all, name = "KZG::commit_variable_batch")] |
| Ok(commitments.into_iter().map(|c| c.into_affine()).collect()) | ||
| } | ||
|
|
||
| #[tracing::instrument(skip_all, name = "KZG::commit_variable_batch_with_mode")] |
There was a problem hiding this comment.
Suggested change
| #[tracing::instrument(skip_all, name = "KZG::commit_variable_batch_with_mode")] | |
| #[tracing::instrument(skip_all, name = "KZG::commit_variable_batch_univariate")] |
| }) | ||
| .collect::<Vec<_>>(); | ||
|
|
||
| // TODO(sagar): Are the coefficients of these witness polynomials always of the same length? |
There was a problem hiding this comment.
can delete this comment
| MultilinearPolynomial::I64Scalars(poly) => { | ||
| Either::Right((i, max_num_bits, poly.coeffs_as_field_elements())) | ||
| } | ||
| _ => Either::Left((i, max_num_bits, poly)), |
There was a problem hiding this comment.
Suggested change
| _ => Either::Left((i, max_num_bits, poly)), | |
| MultilinearPolynomial::U8Scalars(_) => unreachable!("MultilinearPolynomial::U8Scalars cannot have more than 10 bits") |
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.
--features icicleon MacOS (skips the CUDA backend from icicle-jolt)std::BorrowBrings a 1.4x(30%) speedup to the KZG opening proof. GPU opening proofs are ~2.8x(65%) faster than CPU now.
Net improvement of 15% in prover time GPU vs CPU.