Optimum, a startup incubated at MIT and founded by professor Muriel Médard, has launched its private testnet aimed at creating a high-performance memory layer for blockchains. Unlike traditional gossip-based systems, Optimum utilizes Random Linear Network Coding (RLNC) to enhance data propagation efficiency and resilience, achieving up to 20x improvements in bandwidth and double the speed during private tests. The protocol's first product, OptimumP2P, replaces traditional protocols with a publish-subscribe model akin to a memory bus, promising benefits like faster block propagation and reduced congestion. To foster community participation, Optimum will introduce a utility token to reward nodes for their contributions. Although the project currently focuses on refining its memory bus, Médard envisions a decentralized RAM layer with features like atomicity, consistency, and durability coming later this year. Backed by an $11 million seed round and interest from data availability providers, Optimum targets both large and small operators in the blockchain ecosystem. The initiative seeks to address the improvements needed for blockchains to operate as high-speed computing networks.

Source 🔗