Buterin says Ethereum’s biggest bottlenecks are State Tree and VM, and proposes a deep solution

Buterin proposes binary state trees and an eventual RISC-V VM shift to improve Ethereum’s proof efficiency and implementation simplicity.

Vitalik Buterin has proposed changes to the execution layer that could fundamentally reshape Ethereum’s core architecture. The project’s co-founder argued that deep changes to the network’s state tree and virtual machine are needed to remove what he described as the chain’s biggest bottlenecks.

In a detailed post on He called them “basically mandatory” goals if Ethereum wants to enable scalable client-side and zero-knowledge use cases.

Ethereum overhaul

He pointed to EIP-7864, a proposal developed by Guillaume Ballet and others, which would replace Ethereum’s current Keccak-based Merkle Patricia Tree with a binary tree built on a more efficient hash function. According to Buterin, the change would shorten Merkle branches by about four times, by reducing bandwidth requirements and making client-side branch authentication significantly cheaper.

This could reduce data costs for tools such as Helios and private information retrieval systems by a factor of four, Buterin added. Proof efficiency can also be improved three to four times by using only shorter branches. He expects additional gains if Ethereum switches to hash functions like BLAKE3, which are estimated to be three times more efficient than Keccak. Meanwhile, a Poseidon variant could offer up to 100x improvement, although he noted that further security work would be needed.

The proposed binary design would also group storage slots into “pages” of 64-256 slots and enable more efficient loading and editing of adjacent storage, potentially saving more than 10,000 gas per transaction for applications accessing early storage slots. Buterin explained that a prover-friendly state tree would also allow zero-knowledge applications to be constructed directly with Ethereum’s state instead of building independent trees, while simultaneously simplifying the structure and enabling metadata additions for future state decay mechanisms.

In addition to the state tree overhaul, Buterin advocated eventually replacing the Ethereum Virtual Machine with a RISC-V-based VM, describing the idea as longer-term and without consensus. But he expressed strong belief that this would become “the obvious thing” after the state roadmap upgrades are completed.

Possible implementation roadmap

The Ethereum co-founder said a RISC-V VM would be more execution-efficient, prover-friendly and simpler, while noting that many existing provers are already written in RISC-V and an interpreter can be implemented in just a few hundred lines of code. He has laid out a phased transition plan, starting with using the new VM for precompiles, after which developers can deploy contracts directly into the new VM and ultimately retire the EVM into a compatibility layer written as a smart contract in the new system.

You might also like:

Under that roadmap, users would maintain full backward compatibility, barring changes in gas costs, which Buterin said would likely be overshadowed by scaling in the coming years.

Buterin’s latest effort comes just days after he introduced a roadmap for quantum resistance, which included proposals to replace BLS consensus-layer signatures with hash-based schemes such as Winternitz variants.