Skip to content

feat(scheme): add HyperPlonk multi-point to single-point opening reduction#1295

Closed
MavenRain wants to merge 1 commit intoscroll-tech:masterfrom
MavenRain:feat/hyperplonk-open-reduction
Closed

feat(scheme): add HyperPlonk multi-point to single-point opening reduction#1295
MavenRain wants to merge 1 commit intoscroll-tech:masterfrom
MavenRain:feat/hyperplonk-open-reduction

Conversation

@MavenRain
Copy link
Copy Markdown

Toward #904

Phase 1 of #904: standalone sumcheck-based reduction that collapses N = 2^k opening claims f_i(z_i) = y_i (same num_vars) to a single point alpha_2 at which every f_i is opened. Follows HyperPlonk Section 3.8.

Not yet wired into the prover/verifier flow; integration is a follow-up that needs to address heterogeneous num_vars across chips.

…uction

  Phase 1 of scroll-tech#904: standalone sumcheck-based reduction that collapses N = 2^k
  opening claims f_i(z_i) = y_i (same num_vars) to a single point alpha_2
  at which every f_i is opened. Follows HyperPlonk Section 3.8.

  Not yet wired into the prover/verifier flow; integration is a follow-up
  that needs to address heterogeneous num_vars across chips.
@hero78119
Copy link
Copy Markdown
Collaborator

hero78119 commented Apr 6, 2026

Hi @MavenRain thanks for the PR and interests toward this issue

This PR IOP looks quite neat and fit into original issue.
However per recently roadmap change, we plan a more invasively change which aim to enhance original main sumcheck per-chip to main sumcheck for ALL chips. This design make us stay into 1 main sumcheck IOP instead of 1 main sumcheck per chip cascade one multi-point to single opening opening point IOP. I believe your implementation fit latter while we plan to switch to work on former solution.

The newest issue is #1269, and also attach parent issue for reference #1231

@MavenRain
Copy link
Copy Markdown
Author

Hi @MavenRain thanks for the PR and interests toward this issue

This PR IOP looks quite neat and fit into original issue. However per recently roadmap change, we plan a more invasively change which aim to enhance original main sumcheck per-chip to main sumcheck for ALL chips. This design make us stay into 1 main sumcheck IOP instead of 1 main sumcheck per chip cascade one multi-point to single opening opening point IOP. I believe your implementation fit latter while we plan to switch to work on former solution.

The newest issue is #1269, and also attach parent issue for reference #1231

OK, I'll close this since the plan is to move in another direction.

@MavenRain MavenRain closed this Apr 6, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants