AWS Researcher: go-verkle has four major performance bottlenecks and has submitted multiple optimization proposals

ME News Report: On April 2 (UTC+8), Ekleen Kaur, an AWS core protocol researcher, and Everton Fraga conducted high-pressure testing at EthCC[9] on go-verkle, the underlying implementation of the Ethereum Verkle Tree. The testing revealed four major performance bottlenecks blocking the realization of statelessness. 1. During tree updates, a large number of shadow nodes are generated—nodes that do not logically exist but still occupy physical space—leading to roughly 35% redundant expansion of state data; 2. The database index key design includes a large amount of duplicated path information, increasing storage overhead by about 50%; 3. When converting the in-memory tree structure into proofs, there is extensive memory copying rather than pointer references, resulting in deserialization speed 24 times slower than the optimized theoretical value; under high load, this may cause proof verification timeouts; 4. In Verkle tree implementations in clients written in different languages such as Go, Rust, and Nimbus, proofs generated for specific edge cases cannot be recognized across implementations, creating a risk of unexpected forks after upgrades. The two researchers said that AWS has submitted multiple optimization proposals to the Ethereum core repository, focusing on restructuring deserialization logic to eliminate memory copying. After completion, node operation efficiency is expected to improve by more than 3 times, laying the foundation for the formal integration of Verkle trees. (Source: Foresight News)