Ethereum at Hyperscale
summary
Claims a new proof system called Boojum proof system (an implementation of RedShift) and a GPU prover that runs on consumer 6GB GPUs, cutting prover costs by an order of magnitude. Shows recursion that keeps proof size constant and enables effectively unlimited block gas while sequencer throughput remains the bottleneck (current sequencer node in Rust handles north of ~100 TPS, real-world demand ≈10 TPS). Describes volition/zkPorter elastic data availability and a ZK Stack with hyperchains/hyperbridging for low-cost, composable rollups.
AI-written summary · about
chapters (24)
- 0:00Welcome & Introduction with Alex Gluchowski
- 0:38Why zkSync was created
- 1:48Zero-knowledge proofs as a scaling catalyst
- 3:33Capitalism, freedom & permissionless innovation
- 5:11Ethereum as the value layer of the internet
- 6:16What ZK rollups enable: scale and privacy
- 8:25Volition and zkPorter explained
- 10:07zkSync’s cost efficiency and privacy roadmap
- 11:06Recursion and constant-size proof benefits
- 12:05Breakthroughs and proof system: Boojum
- 14:22Performance benchmarks and GPU proving
- 16:57Mainnet rollout and shadow proving
- 17:56Unlimited block size with ZK recursion
- 19:03zkSync throughput and battle-tested usage
- 20:12ZK vs Optimistic rollups
- 23:28State diffs vs calldata and compression
- 25:10Volition trade-offs and user security
- 26:45ZK Stack and EVM compatibility
- 28:41Modular deployments with hyperchains
- 30:23Hyperchains as city districts
- 33:38Hyperbridging and seamless interoperability
- 35:18Lessons learned and early missteps
- 36:55How to measure Ethereum scalability success
- 38:38ZK Credo and defining hyperscalability
related talks
25mSuccinct Proofs in Ethereum
36mJoin Sandeep Nailwal, Rene Reinsberg, Alex Gluchowski & Steven Goldfeder I Pragma Denver 2024
16mProving Ethereum Full Nodes in ZK
16mTruth in a Post Truth World
19mThe Superpowers of the ZK Stack
24mZK: Accelerating DA Interop
31mNext 10 Years of Ethereum
20mProvers Networks from First Principle