BlockBeats News, September 9 — Ethereum co-founder Vitalik Buterin has published a proposal introducing the recursive STARK mempool proposal EIP-8288, aiming to include it in the I-star upgrade following Hegota. The proposal moves signatures and proofs outside Ethereum's core execution path and recursively aggregates STARKs in the mempool, reducing on-chain computation and data costs.
Vitalik stated that this approach could support low-cost quantum-safe signatures and privacy protocols. Specifically, the cost of quantum-safe privacy transactions is expected to drop from approximately 10 million Gas to tens of thousands of Gas, and it can accommodate new signature or proof schemes such as Falcon and ML-DSA without modifying the EVM, while also supporting private account abstraction.
By design, nodes will periodically aggregate transaction dependencies and generate recursive STARKs, with block builders then generating proofs for transactions intended for inclusion in a block. The additional on-chain overhead per block is roughly one STARK of 100 to 300 KB, plus 96 bytes of data per statement to be proven. This proposal could also drive Ethereum to establish RISC-V as the standard instruction set for recursive STARKs.

