zkSaaS: Zero-Knowledge SNARKs as a Service
summary
Builds a distributed protocol that secret-shares a witness across one large host and many commodity workers, runs an MPC in the exponent to avoid blowing up crypto ops, and splits proof generation into FFTs, multiscalar multiplications and sum-of-partial-products subroutines with constant‑round low‑communication MPCs. Implementation on Groth-16 reduces prover time by about an order of magnitude, targets per-node work ≈1/n, and achieves security against < n/4 corrupted parties.
AI-written summary · about
chapters (8)
- 0:00Introduction and context
- 0:38Introduction to ZK-SAS
- 1:12Understanding ZK-SNARKs
- 5:21Applications and growth
- 7:27Computational bottlenecks
- 9:08Existing solutions and trade-offs
- 13:10Proposed framework for ZK-SAS
- 15:46Implementation and results
related talks
19mDIZK: A Distributed Zero Knowledge Proof System
23mSonic: Nearly Trustless SNARK Setups
56mCollaborative zk-SNARKs: Proving as One over Distributed Secrets
25mPractical Accountability of Secret Processes
48mzkSNARKs — Cryptographic Proofs for Scaling and Security Challenges
23mZKBoo: Faster Zero-Knowledge for Boolean Circuits
33mAn Introduction to Zero Knowledge Proofs
39mPrivacy and Evolution of SNARKs