Brevis CEO outlines ZK-powered 'infinite computing layer' for blockchain

During a CoinNess Night Live event on Jan. 29, Brevis CEO and co-founder Michael explained that the project is a verifiable computing platform based on zero-knowledge (ZK) proofs. It processes complex, costly, and slow computations off-chain, then uses a small cryptographic proof to verify the results on-chain. Michael described the project's "infinite computing layer" as a structure where the blockchain only verifies proofs rather than re-executing computations. This model allows for virtually unlimited off-chain scaling of high-cost operations like transaction history analysis, AI inference, and cross-chain verification. He noted that Brevis technology is already being used for functions such as volume-based fee discounts and reward distribution, with users experiencing the benefits seamlessly as all calculations are handled off-chain and verified with ZK proofs. Key products from Brevis include: - Pico: A Rust-based modular zkVM that proves Ethereum blocks in real-time. - ZK Data Coprocessor: Enables smart contracts to trustlessly access historical on-chain data. - ProverNet: A decentralized marketplace for ZK proof generation, currently live on mainnet. - Incentra: A trust-minimized platform for incentive distribution using off-chain computation. Regarding its role in the Ethereum Foundation's roadmap, Michael stated that Pico is already operational within the Ethproofs infrastructure. Brevis is working to enable faster proofs on lighter hardware, aiming for a future where anyone can participate in Ethereum validation from a mobile device. He also revealed that Brevis is testing its Pico Prism in a 16 GPU environment, down from the previous 64 GPUs, to make Ethereum block proofs cheaper and more accessible. The project is also preparing to integrate new functionalities related to privacy and AI verification.