Vitalik Buterin lays out a two-part plan to overhaul Ethereum’s execution layer from the ground up

Ethereum co-founder Vitalik Buterin published a detailed breakdown of two major execution layer changes he sees as critical to the network’s future, doubling down on a binary state tree overhaul and a longer-term push to move beyond the EVM.

In an X post Sunday, Buterin argued the two changes are “deep” architectural shifts that many developers shy away from, but that incrementalism will not resolve Ethereum’s core proving inefficiencies. The state tree and VM together represent more than 80% of the proving bottleneck, he said, making both “basically mandatory” for client-side proving use cases.

“They are ‘deep’ changes that many shrink away from, thinking that it is more ‘pragmatic’ to be incrementalist,” Buterin noted.

State trees

The state tree change centers on EIP-7864, which would replace Ethereum’s current hexary Keccak Merkle Patricia Tree with a binary tree using a more efficient hash function. Buterin credited developer Guillaume Ballet and others for their work on the proposal, which has been in draft stages since January 2025.

The binary tree design would produce Merkle branches four times shorter than today’s structure, cutting data bandwidth for light client tools like Helios. On top of that, swapping the hash function to either Blake3 or a Poseidon variant could deliver an additional 3x to 100x improvement in proving efficiency, though Buterin noted that Poseidon still requires more security review.

The binary tree shift represents a notable evolution from earlier plans. Verkle Trees were previously cited as a leading candidate for inclusion in a 2026 hard fork, but concerns around quantum computing vulnerabilities in elliptic curve cryptography helped reignite interest in binary trees around mid-2024.

A new VM?

On the VM front, Buterin reiterated his April 2025 proposal to eventually replace the EVM with RISC-V, the open-source instruction set architecture that most ZK provers already use internally. He described a three-stage deployment: first using RISC-V only for precompiles, then allowing users to deploy RISC-V contracts, and finally retiring the EVM entirely by converting it into a smart contract written in the new VM.

“Ethereum’s whole point is its generality, and if the EVM is not good enough to actually meet the needs of that generality, then we should tackle the problem head-on, and make a better VM,” Buterin wrote.

The RISC-V proposal has not gone unchallenged. In November 2025, researchers from Offchain Labs, the core developer behind Arbitrum, published a detailed rebuttal arguing that WebAssembly (WASM) is a better long-term choice for Ethereum’s smart contract format. Their core point: RISC-V excels at ZK proving, but the “delivery ISA” and the “proving ISA” do not need to be the same.

The proposals come as Buterin has become an increasingly vocal champion of deep structural change to Ethereum’s base layer. Last week, he said Ethereum “has already made jet engine changes in-flight once (the merge)” and can handle roughly four more: the state tree, lean consensus, ZK-EVM verification, and a VM change.

Ethereum’s Glamsterdam upgrade is expected in the first half of 2026, with the Hegota upgrade to follow later in the year. Developers have not finalized the headline EIP for either fork, but state tree changes and execution layer improvements remain central to ongoing planning.