Episode Summary
Executive Summary: The episode explores how Type 1 ZK EVMs can “snarkify” Ethereum, meaning they can prove full EVM execution with succinct cryptographic proofs while preserving the exact Ethereum spec. Justin Drake frames this as key to enshrined rollups, lighter validators, higher L1 throughput, and eventually multi-core Ethereum. Brian Redford explains Risk Zero’s Zeth and Bonsai as practical infrastructure to prove Ethereum blocks, optimize proof generation, and extend ZK to rollups and apps.
Main Topics: Snarkifying Ethereum (Priority: 5/5): Justin explains the roadmap to prove both the EVM and beacon chain with ZK proofs so computation stops being a consensus bottleneck and Ethereum can scale with cryptography rather than raw hardware. Type 1 ZK EVMs and enshrined rollups (Priority: 5/5): Brian describes Zeth as a Type 1 ZK EVM that matches Ethereum’s specification exactly, making it a candidate for enshrined rollups, unlike modified Type 2 systems that compromise on compatibility. Performance and proving architecture (Priority: 4/5): The discussion covers how Risk Zero uses RISC-V, Rust, GPUs, and recursive/continuation-based proving to cut proof generation time and enable eventual low-latency validation. Security, auditability, and trust assumptions (Priority: 4/5): The speakers discuss the security surface of ZK systems, the benefits of a smaller instruction set, and the new risks introduced by compilers, databases, and hardware acceleration. Bonsai as ZK infrastructure (Priority: 4/5): Risk Zero positions Bonsai as a ZK application development platform and co-processor network for rollups, DeFi, and eventually gaming, enabling off-chain computation with on-chain attestations. Rollup ecosystem and interoperability (Priority: 4/5): They discuss a partnership path with Optimism to ZKify the OP Stack, showing how ZK proofs may accelerate liquidity movement, fraud-proof replacement, and faster L2 interoperability. Ethereum’s long-term decentralization and UX (Priority: 5/5): The conversation envisions validators on phones or smartwatches, tiny bandwidth requirements, and Ethereum becoming as easy to access as HTTPS-secured web pages.
Key Arguments: Type 1 ZK EVMs preserve full Ethereum compatibility, which is essential if Ethereum wants an enshrined rollup inside Layer 1. Once the EVM is snarkified, block verification can become millisecond-fast, enabling much higher gas limits and eventually more L1 throughput. ZK proofs reduce the need for expensive full-node hardware, lowering the barrier for validators and users and strengthening decentralization. Data availability sampling and ZK execution together attack the two main consensus bottlenecks: computation and bandwidth. Risk Zero’s approach leverages existing ecosystems—Rust, LLVM, and RISC-V—to take a major shortcut from ordinary software to provable Ethereum execution. Security is improved by shrinking the audited ZK surface, but new trust assumptions arise in compilers and infrastructure, so audits and formal methods remain critical. The near-term market for this technology is rollups, L2s, L3s, and ZK co-processing for applications; Ethereum mainnet enshrined rollups are the longer-term endgame.
Data Points: OP grant for ZKifying the OP Stack: 250,000 OP tokens - Justin cites the Optimism RFP funding level for teams to snarkify Optimism Approximate grant value: $375,000 - Justin estimates the dollar value of the OP token grant at current prices Current Ethereum throughput: ~16 transactions per second - Justin references rough mainnet throughput before scaling via ZK and rollups Combined rollup throughput mentioned: ~5x Ethereum base throughput - The hosts cite L2B-style combined rollup activity relative to Ethereum L1 Proof latency target for validator use cases: ~1 second - Justin says blocks should be verifiable within about a second for validator-grade UX Current proof latency for the system: ~50 minutes with 64 machines - Brian describes the current proof generation time on off-the-shelf hardware Improved proof latency with more machines: ~12 minutes - Brian says more parallel hardware can reduce proof time substantially Distance from target latency: 100x to 1000x away - Justin characterizes current performance relative to the one-second goal Hardware used for proving: CPUs, M2 GPU, NVIDIA GPUs - Brian lists supported proving targets Performance improvement thesis: ~4x per year - Brian describes rapid ZK progress as resembling a Moore’s-law-like trend Recursion depth: 10 vertical steps - Brian explains the recursion stack needed to roll 1024 proofs into one Block chunking reference: 1024 proofs - Used as an example of splitting computation for recursive proof aggregation
Pivotal Quotes: "We snarkify everything." — Justin Drake: Used to describe the Ethereum endgame where core protocol components are proven with ZK "The future is very bright given all of the advances that are being made across the entire space." — Brian Redford: Brian’s reaction to the pace of cryptography and Ethereum infrastructure development "Once we’ve snarkified the canonical instance of the EVM that we have today, we’re going to have one enshrined roll-up." — Justin Drake: Justin explains the relationship between Type 1 ZK EVMs and enshrined rollups
Implications: ZK proofs are moving from theory to infrastructure, with major implications for rollups, Ethereum UX, validator decentralization, and L1 scaling. Expect near-term adoption in L2s and app-level co-processing, with enshrined rollups and ultra-light clients as the long-term prize.