Replacing the EVM with RISC-V to Boost Ethereum Scalability

Vitalik Buterin has introduced a proposal to fundamentally rework Ethereum's execution environment by transitioning from the current Ethereum Virtual Machine (EVM) to RISC-V

In a surprising move that could fundamentally rework Ethereum’s execution environment, Vitalik Buterin has proposed transitioning from the current Ethereum Virtual Machine (EVM) to RISC-V.

The proposal, published on April 20 in the Ethereum Magicians forum, suggests a multi-phase strategy to massively improve long-term proving efficiency and simplify ETH’s execution layer—all while maintaining core abstractions like accounts, storage, and contract calls.

Solidity and Vyper Stay, Rust Optional

Specifically, Buterin proposes keeping Solidity and Vyper as the primary smart contract languages, but having them adapted to compile directly to RISC-V instructions.

Although Rust-based contract writing would be technically feasible, Buterin argues that developer familiarity and readability concerns make Solidity the more logical choice at the application layer.

Importantly, the upgrade would be fully backward compatible. Existing EVM-based contracts would continue to function and interact seamlessly with contracts compiled to RISC-V, ensuring a smooth transition across the network.

Targeting Ethereum’s Final Bottleneck

According to Buterin, execution is shaping up to be Ethereum’s last major long-term bottleneck as other scalability improvements are handled through upgrades like:

* Merkleization and efficient state transitions for efficient state proofs.

* ZK rollups and efficient parallelism.

The main pain point is in ZK-EVM proving costs. Recent analysis from Succinct Labs shows that block execution consumes nearly 50% of total prover cycles in ZK rollups. While efforts are being made to reduce state tree overhead by moving from Keccak-based Merkle-Patricia tries to binary tries with hash functions like Poseidon, execution is the dominant constraint.

Buterin noted that ZK-EVMs already compile to RISC-V at a lower level, and exposing RISC-V directly as Ethereum’s virtual machine could remove a major translation overhead. Some tests suggest this could yield up to 100x improvements in prover performance.

"We should think in decades," Buterin said. "This generation of ZK rollups is already limited by prover performance, and making Ethereum's execution layer future-proof is a step worth taking."

If adopted, this proposal could mark one of Ethereum’s most transformative infrastructure upgrades yet—bridging today’s high-level developer tools with next-gen zero-knowledge performance.