Ethereum co-founder Vitalik Buterin proposes replacing the EVM with RISC-V instruction set architecture

According to Buterin, the aim is to dramatically enhance the efficiency, simplicity, and scalability of Ethereum's execution layer.

Ethereum co-founder Vitalik Buterin has proposed a major change to the foundation of the network's execution environment: replacing the long-standing Ethereum Virtual Machine (EVM) with the open-source RISC-V instruction set architecture.

The aim is to dramatically enhance the efficiency, simplicity, and scalability of Ethereum's execution layer, Buterin said in a post on Wednesday on Ethereum's core developer forum.

The EVM has been at the heart of Ethereum's smart contract infrastructure, serving as a key component not only for the blockchain itself but also for countless other blockchains and Layer-2 rollups built to be EVM-compatible.

However, Buterin argues that the EVM may be reaching its architectural limits. A switch to RISC-V could yield efficiency gains of up to 100x for certain operations, especially in zero-knowledge proof generation, he added.

Unlike the EVM, RISC-V is a standardized, hardware-friendly architecture designed to be simple, extensible, and efficient. Its adoption in Ethereum could reduce execution layer complexity, improve performance in zkEVM proving, and enhance support for cryptographic operations thanks to its native instruction support.

This would also not mean abandoning Solidity or existing EVM contracts. Buterin envisions a dual-VM strategy, allowing Solidity code to compile down to RISC-V bytecode.

There are also discussions around maintaining backwards compatibility by running both EVM and RISC-V contracts side by side or creating a RISC-V-based interpreter for legacy contracts.

Buterin's proposal comes at a time when ETH is grappling with multiple challenges. Transaction fees have dropped to multi-year lows, network activity has declined, and rival chains are gaining ground in performance and developer adoption.

With many users and developers shifting to Layer-2s, concerns are also rising over Ethereum's long-term revenue generation and market relevance.

While Ethereum's consensus layer continues to evolve through initiatives like the Beam Chain and other research projects, the execution layer remains a bottleneck. Buterin believes that only a "radical change" like RISC-V could unlock the next level of efficiency and scalability.

"We can get to a parallel RISC-V chain that has about 100x throughput for certain types of operations (especially useful for zkEVM proving), is about 5-10x smaller code size, and has excellent support for cryptographic primitives," Buterin said.

"It's not an impossibility to get to a chain that does both at the same time, and crucially, has a modular design that allows for pieces like the crypto primitives to be easily modified or upgraded."