EVM should exit the stage! Vitalik proposes switching to RISC-V, ZK proof performance can be improved by more than a hundred times.

Ethereum co-founder Vitalik yesterday made a major proposal to replace the current EVM language with RISC-V, an open-source instruction set architecture, to simplify the execution layer and improve the efficiency of smart contracts and ZK proofs. (Synopsis: Abandoning miners to become a failure of ETH? Analysis: After Ethereum switches to PoS in 2022, "funding attention is getting lower and lower") (Background addition: Ethereum's handling fees hit a five-year low, and the community set off the "L2 poison theory": there are no cars on the road, and V God is still laughing and building highways) At a time when Ethereum faced competitive pressure from other public chains and the Layer2 strategy failed, co-founder Vitalik Buterin threw out another "bold reform": replacing the current Ethereum Virtual Machine (EVM) with the RISC-V instruction set architecture. It attempts to replace the existing architecture with a more concise and common underlying system and reshape the efficiency of smart contract execution. Vitalik Throws Shock Bombs: EVM Unplugged for RISC-V? Vitalik published a proposal at the Ethereum Magicians forum yesterday (20), pointing out that for a long time, Ethereum's scaling and performance have always been limited by execution layer bottlenecks, especially when combined with ZK-EVM (zero-knowledge virtual machine), proving that the cost remains high. He argues that changing the underlying language for executing smart contracts from EVM to RISC-V will make contract execution faster, the system simpler, and better expanded in the future. The so-called RISC-V is an open-source reduced instruction set architecture, compared with the current Ethereum-exclusive virtual machine language, RISC-V is closer to the underlying language of the traditional CPU, and the development tools and support environment are more extensive. Vitalik says most developers can still develop in languages like Solidity or Vyper by simply compiling the backend to RISC-V. RISC-V proof efficiency increased by more than 100 times Citing ZK-EVM's cycle data*, Vitalik pointed out that the four most resource-consuming steps of the execution layer at present include: input data decoding (deserialization), state verification (initializing the witness database), block execution and state root calculation, of which the most resource-consuming steps are block execution and input data decoding, which together account for more than half of the total computation. Note: The so-called "cycle data" refers to the processing cycle consumed by each execution step when generating ZK proofs, which can be regarded as a quantitative indicator of resource consumption. Taking the cycle count displayed in the chart as an example, different block types (such as blocks with bn254 or bls12_381 syscall) alone account for 31%~60% of the execution cost. He pointed out that many ZK-EVMs themselves have compiled the EVM to RISC-V and then performed zero-knowledge proofs, so "it is actually more efficient for developers to use RISC-V directly." Vitalik believes that if the contract is run directly on RISC-V, the overall ZK proof performance improvement of more than 100 times is not a dream. According to data provided by Viglik, EVM implementations can perform intensive calculations such as Fibonacci up to 800 times slower than native RISC-V. It is worth noting that Vitalik's proposal is not a one-time replacement, but supports a "dual VM coexistence" model: the EVM contract and the RISC-V contract can coexist and call each other without disrupting the existing smart contract operation. In addition, he proposed a further vision of a "virtual machine interpreter" that runs traditional EVM contracts through interpreters written in RISC-V for a smooth transition, further unifying the underlying language and greatly simplifying execution layer specifications. HE EMPHASIZED THAT ETHEREUM'S PAST HAS BEEN DIFFICULT TO MAKE ANY MINOR SIMPLIFICATIONS AT THE PROTOCOL LAYER (SUCH AS REMOVING SELFDESTRUCT), AND THAT ADOPTING RISC-V MAY BE THE ONLY PRAGMATIC WAY TO CREATE A LONG-MAINTAINABLE, CONCISE EXECUTION LAYER. Ethereum fees hit a five-year low The proposal comes at a time when the Ethereum mainnet is under economic pressure. On-chain data shows that the average transaction fee of Ethereum fell to $0.16 in April, the lowest since 2020, indicating that mainnet transaction demand has decreased significantly, and more and more users are turning to Layer 2, resulting in dilution of mainchain revenue. The price of Ether (ETH) also fell into the doldrums, trading around $1,600. Many analysts warn that if the main chain of Ethereum cannot recover its revenue and scaling advantages, the price of ETH may continue to fall. Can RISC-V reverse Ethereum's decline? Vitalik's proposal is not only a technical architecture reform, but also a turning point for Ethereum to rebuild market confidence. When a new generation of high-throughput public chains such as Solana and Sui seize the attention of developers and funds with a more concise monolithic architecture, if Ethereum can greatly improve execution efficiency without sacrificing security and decentralization, it will redefine the competitive threshold of L1. Although this architectural change is still in the conception stage, if it is implemented, it will affect the development tool chain, node operating model and layer 2 integration strategy, which is one of the most systematic changes in Ethereum in a decade. Related reports A single family is so big that it makes enemies on all sides, is Ethereum okay? Ethereum faces generational calls! V God: The growth of the application layer needs more "good social philosophy", naming the advantages and disadvantages of the application Xiao Feng Hong Kong Dialogue Vitalik: One year to learn Chinese, Ethereum Foundation opens a branch in Hong Kong: 〈It's time for EVM to quit! Vitalik's proposal to switch to RISC-V, ZK proves that the performance can be improved by more than 100 times" This article was first published in BlockTempo's "Dynamic Trend - The Most Influential Blockchain News Media".