Blockchain Performance, Demystified

This podcast interviews Aptos Labs' Head of Research, Sasha Spiegelman, and a16z crypto head of research, Tim Roughgarden, about blockchain performance. The discussion centers on key performance metrics (transactions per second and latency), highlighting the challenges in comparing different blockchains due to varying definitions. Spiegelman details Aptos' innovations, including a novel approach to consensus that decouples data dissemination from consensus logic, resulting in a 10x increase in throughput. The conversation also explores the trade-offs between optimistic and pessimistic approaches to parallel transaction execution and the ongoing debate about dynamic versus static parallelism in blockchain design. The interview concludes with a discussion on the importance of open-source collaboration in advancing blockchain technology.