Ethereum co-founder Vitalik Buterin has laid out a detailed and highly technical case for two sweeping execution layer changes that aim to rebuild the network’s core from the ground up. In a move that shifts the scaling conversation away from Layer 2 rollups, Buterin proposed a transition to a “Binary State Tree” and the eventual replacement of the Ethereum Virtual Machine (EVM) with a RISC-V architecture.
The Proving Bottleneck
For years, the Ethereum community has focused heavily on scaling through Layer 2 solutions. However, Buterin recently argued that incrementalism will not resolve the base layer’s fundamental inefficiencies. According to his analysis, the network’s state tree and virtual machine together account for more than 80% of Ethereum’s proving bottleneck. Addressing these core architectural constraints has become practically mandatory for the network’s long-term zero-knowledge (ZK) scalability goals.
These proposed upgrades represent deep structural changes that many developers have historically shied away from due to their complexity. Yet, as ZK technology becomes central to Ethereum’s roadmap, the base layer itself must evolve to become more “prover-friendly.” The goal is to allow ZK applications to integrate directly with Ethereum’s state efficiently, rather than forcing developers to build parallel, fragmented systems.
“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. They are ‘deep’ changes that many shrink away from, thinking that it is more ‘pragmatic’ to be incrementalist.”
Vitalik Buterin, Ethereum Co-founder
The Binary State Tree Transition
The first major pillar of this overhaul centers on EIP-7864, which proposes replacing Ethereum’s current hexary Keccak Merkle Patricia Tree with a leaner binary tree structure. This represents a significant pivot from earlier plans, where Verkle Trees were the leading candidate for the 2026 hard fork. Concerns around the long-term cryptographic risks of elliptic curve cryptography in a post-quantum world ultimately reignited interest in the binary tree design.
The technical benefits of this shift are profound. A binary tree design produces Merkle proofs that are roughly four times shorter than the current structure. This dramatic reduction in data bandwidth makes lightweight clients—like Helios—significantly cheaper and more viable. Furthermore, the new structure would group storage slots into “pages,” allowing decentralized applications that repeatedly access adjacent data to operate with unprecedented efficiency.
| Architecture Metric | Current Merkle Patricia Tree | Proposed Binary State Tree |
|---|---|---|
| Proof Size | Large (High bandwidth) | 4x Shorter (Low bandwidth) |
| Storage Access | Individual slots | Grouped “pages” for efficiency |
| Gas Savings | Baseline | Up to 10,000+ gas saved per transaction |
| Quantum Resistance | Vulnerable | Highly adaptable (Blake3/Poseidon) |
Retiring the EVM for RISC-V
The most radical aspect of Buterin’s proposal is the long-term vision to phase out the Ethereum Virtual Machine (EVM) entirely. He advocates for a transition to RISC-V, a widely used, open-source instruction set architecture that most ZK provers already utilize internally. The increasing reliance on special-case precompiles in Ethereum reflects a growing discomfort with the EVM’s inherent limitations when handling complex cryptographic operations.
The transition would not happen overnight. Buterin outlined a meticulous three-stage deployment plan. First, RISC-V would be used exclusively for precompiles, acting as a “GPU for the EVM” to accelerate math-heavy operations. Second, users would be allowed to deploy RISC-V contracts directly. Finally, the EVM would be retired entirely by converting it into a legacy smart contract written within the new RISC-V virtual machine, preserving backward compatibility while unlocking raw execution efficiency.
The Architectural Debate
This aggressive push to rewrite Ethereum’s foundation has not gone unchallenged. The tension reflects a broader debate over whether Ethereum should continue layering solutions on top of its existing design or fundamentally rework its core engine. Supporters argue that aligning the base layer’s architecture with modern ZK proving systems is the only viable path to true global scale without compromising decentralization.
Conversely, critics and rival developers have pushed back. Researchers from Offchain Labs, the team behind Arbitrum, recently published a detailed rebuttal arguing that WebAssembly (WASM) is a better long-term choice for Ethereum’s smart contract format than RISC-V. Other analysts warn against the growing abstraction across the Ethereum roadmap, arguing that each deep architectural change increases complexity, introduces new trust assumptions, and creates fresh attack surfaces during the multi-year transition period.
Final Thoughts
Vitalik Buterin’s proposed overhaul of the execution layer serves as a technical turning point for Ethereum. By targeting the state tree and the virtual machine, the network is preparing to address the very bottlenecks that threaten its long-term viability as a universal settlement layer.
The divergence between the pragmatic desire for incremental upgrades and the visionary push for a foundational rewrite is striking. As developers prepare for the Glamsterdam upgrade in 2026, the question remains: can Ethereum successfully execute a “jet engine change in-flight” for a second time, or will the sheer complexity of replacing the EVM fracture the consensus that holds the network together?
