After integrating Reth 2.5.1, BSC-Reth achieved about 40% higher throughput than the TrieDB-based configuration—according to BNB Chain’s internal benchmarks. The migration goal is to move validators to the BSC-Reth client by the end of 2026.
In the same tests, block write latency (p95) was 71.6 milliseconds, 34.9% lower than BSC-Geth; execution latency (p95) was 56.5 milliseconds, down 35.5%. These figures compare client-level performance; they are not measurements from actual mainnet operations.
· Support for BEP-675, which is intended to extend block-building time, is still under evaluation. Block-level access lists (BALs) and EVM2 are also only being explored.
Which would you rather see first: real-world BSC-Reth latency measurements on mainnet, or a specific timeline for the validator migration?
In the same tests, block write latency (p95) was 71.6 milliseconds, 34.9% lower than BSC-Geth; execution latency (p95) was 56.5 milliseconds, down 35.5%. These figures compare client-level performance; they are not measurements from actual mainnet operations.
· Support for BEP-675, which is intended to extend block-building time, is still under evaluation. Block-level access lists (BALs) and EVM2 are also only being explored.
Which would you rather see first: real-world BSC-Reth latency measurements on mainnet, or a specific timeline for the validator migration?
