Ethereum co-founder Vitalik Buterin has unveiled a comprehensive proposal to fundamentally restructure Ethereum's core architecture to address scaling challenges. At the center of Buterin's proposal is EIP-7864, which would replace Ethereum's current hexary Merkle Patricia tree with a simpler binary tree design. According to reports from Yellow.com, the change would produce Merkle proofs roughly four times shorter than the existing structure, cutting verification bandwidth and making lightweight clients far cheaper to run. Buterin described these as 'the big bottlenecks that we have to address if we want efficient proving', marking a significant shift in how Ethereum's leadership frames the scaling problem. The new binary tree would also group storage slots into pages, allowing decentralized applications that load related data to do so more efficiently — potentially saving more than 10,000 gas per transaction in some cases.
Beyond the state tree changes, Buterin outlined a longer-term vision to move past the Ethereum Virtual Machine altogether in favor of a RISC-V-based architecture. As reported by Yellow.com, he described a phased deployment approach: RISC-V would first power precompiles, then support user-deployed contracts, and eventually absorb the EVM itself as a compatibility layer. In the near term, he proposed a 'vectorized math precompile' — what he called a 'GPU for the EVM' — to accelerate cryptographic operations. This represents a fundamental shift from traditional Layer 2 rollup solutions to deep-layer protocol changes that would enable zero-knowledge applications to integrate directly with Ethereum's base layer instead of building parallel systems.
Buterin's state tree proposal comes alongside his 60% probability that SNARKs, FHE, and iO will reach sub-10x overhead - the threshold he frames as practical for mainstream blockchain applications. According to reports from CoinDesk, Buterin's estimate, posted to X in 2026, reflects his evolving assessment of how quickly the field's most powerful privacy and scaling tools can shed the computational burden that has kept them confined to research environments. The 60% figure is not tied to a specific calendar deadline, functioning as a directional bet rather than a dated milestone. In a September 6 analysis, Buterin went even further, putting a 33% chance on those same technologies reaching near-zero overhead (expressed mathematically as 1+ε) at large scales, meaning there's roughly a one-in-three shot that running encrypted computations could become almost as cheap as running them in the clear.
Not all industry participants agree with the proposed deep-layer changes. Analyst DBCrypto criticized what he described as growing abstraction across the Ethereum roadmap, arguing that each additional layer increases complexity, introduces trust assumptions and creates potential attack surfaces. As reported by Yellow.com, the debate now centers on whether the proposed foundational rework is worth the complexity it introduces. The proposal marks a significant departure from traditional scaling approaches, focusing on deep protocol changes rather than Layer 2 solutions or blob capacity improvements. This shift reflects Buterin's belief that zero-knowledge applications could integrate directly with Ethereum's base layer if the underlying architecture becomes prover-friendly, potentially revolutionizing how privacy-preserving applications operate on the network.
