In a bold move, Vitalik Buterin is redirecting the Ethereum scaling discussion back to the core protocol, emphasizing that the primary long-term limitations are not found in Layer 2 solutions or blob capacity, but rather in deeper architectural bottlenecks within the network”s state tree and virtual machine.
Buterin”s insights reveal that these two components—the state tree and the virtual machine—are responsible for over 80% of the proving costs associated with operations on the Ethereum network. This is particularly crucial as zero-knowledge (ZK) technology takes center stage in Ethereum”s future roadmap.
The centerpiece of his proposal is EIP-7864, which seeks to replace Ethereum”s existing hexary Merkle Patricia tree with a more efficient binary tree structure. While this alteration may appear minor, its potential impact is substantial. The binary tree design could yield Merkle proofs that are approximately four times shorter than those produced by the current system, significantly lowering verification bandwidth needs. This would make lightweight clients and privacy-focused applications both cheaper and more feasible.
Furthermore, the new structure will organize storage slots into “pages,” enabling applications that frequently access related data to do so more effectively. Many decentralized applications (dApps) often need to access adjacent storage slots, and this improvement could lead to gas savings exceeding 10,000 per transaction in certain scenarios.
Buterin has also proposed coupling the tree redesign with more effective hash functions, which could further enhance proof generation speeds. This reconfiguration would ultimately create a more “prover-friendly” base layer for Ethereum, allowing ZK applications to integrate directly with Ethereum”s state rather than relying on parallel systems.
On a broader scale, the binary tree initiative aims to encapsulate a decade”s worth of lessons in state management, transitioning to a more streamlined and future-ready architecture.
In addition, Buterin shared his long-term vision for Ethereum”s execution engine, suggesting a potential shift from the Ethereum Virtual Machine (EVM) to a RISC-V-based architecture. RISC-V is a widely adopted open instruction set that could provide improved efficiency and simplicity. Buterin noted that Ethereum”s growing dependence on specialized precompiles indicates a discomfort with the EVM”s limitations. He posits that if Ethereum”s fundamental promise is general-purpose programmability, the virtual machine should fully realize that vision without unnecessary workarounds.
He envisions that a RISC-V-based virtual machine could offer several advantages, including the introduction of a “vectorized math precompile,” likened to a “GPU for the EVM,” aimed at significantly speeding up cryptographic operations. Over time, he outlined a phased transition where RISC-V would first support precompiles, subsequently enable user-deployed contracts, and ultimately replace the EVM itself as a compatibility layer.
However, there are voices of skepticism regarding the need for such deep-layer modifications. Analyst DBCrypto has criticized what he perceives as an increasing abstraction within the Ethereum roadmap, which includes new frameworks designed to tackle rollup fragmentation. He argues that each additional layer complicates the architecture, introduces new trust assumptions, and creates additional potential vulnerabilities.
This ongoing debate highlights a significant question: should Ethereum continue to build on its existing framework or reconsider its foundational design? But according to Buterin, the architecture must evolve to accommodate the rising importance of zero-knowledge proofs, suggesting that the next phase of scaling may not be confined to Layer 2 but instead take place deep within Ethereum”s core.
