{"id":94822,"date":"2026-03-02T10:23:48","date_gmt":"2026-03-02T07:23:48","guid":{"rendered":"https:\/\/forklog.com\/en\/?p=94822"},"modified":"2026-03-03T10:26:19","modified_gmt":"2026-03-03T07:26:19","slug":"buterin-proposes-a-binary-state-tree-and-a-long-term-replacement-for-the-evm-in-ethereum","status":"publish","type":"post","link":"https:\/\/forklog.com\/en\/buterin-proposes-a-binary-state-tree-and-a-long-term-replacement-for-the-evm-in-ethereum\/","title":{"rendered":"Buterin proposes a binary state tree and a long-term replacement for the EVM in Ethereum"},"content":{"rendered":"<p>Ethereum co-founder Vitalik Buterin set out a plan for two key changes to the <span data-descr=\"the part of the architecture responsible for processing transactions and executing smart-contract code via the EVM\" class=\"old_tooltip\">network execution layer<\/span>: a shift to a binary state tree and the long-term replacement of the <a href=\"https:\/\/forklog.com\/en\/news\/what-is-the-ethereum-virtual-machine-evm\">EVM<\/a>.<\/p>\n<blockquote class=\"twitter-tweet\">\n<p lang=\"en\" dir=\"ltr\">Now, execution layer changes. I&#8217;ve already talked about account abstraction, multidimensional gas, BALs, and ZK-EVMs.<\/p>\n<p>I&#8217;ve also talked here about a short-term EVM upgrade that I think will be super-valuable: a vectorized math precompile (basically, do 32-bit or potentially\u2026<\/p>\n<p>\u2014 vitalik.eth (@VitalikButerin) <a href=\"https:\/\/twitter.com\/VitalikButerin\/status\/2028158949720252574?ref_src=twsrc%5Etfw\">March 1, 2026<\/a><\/p><\/blockquote>\n<p> <script async src=\"https:\/\/platform.twitter.com\/widgets.js\" charset=\"utf-8\"><\/script><\/p>\n<blockquote class=\"wp-block-quote is-layout-flow wp-block-quote-is-layout-flow\">\n<p><em>&#8220;These are &#8216;deep&#8217; changes that many forgo, considering the <span data-descr=\"a method of project management and product development in which work is done not all at once but in increments\" class=\"old_tooltip\">incremental approach<\/span> more &#8216;pragmatic&#8217;,&#8221; he noted.<\/em><\/p>\n<\/blockquote>\n<p>According to the programmer, the state tree and the virtual machine account for over 80% of the network\u2019s technical constraints. This makes them &#8220;practically mandatory&#8221; for client scenarios that use proofs.<\/p>\n<p>According to the programmer, the state tree and the virtual machine account for over 80% of the network\u2019s technical constraints. This makes them &#8220;practically mandatory&#8221; for client scenarios that use proofs.<\/p>\n<h2 class=\"wp-block-heading\">A new state tree<\/h2>\n<p>The state-tree change builds on <a href=\"https:\/\/eips.ethereum.org\/EIPS\/eip-7864\">EIP-7864<\/a>. It entails moving from the current hexadecimal <a href=\"https:\/\/forklog.com\/en\/news\/what-is-a-merkle-tree\">Merkle\u2013Patricia tree<\/a> to a binary one with a more efficient hash function.<\/p>\n<p>A binary structure would shorten Merkle branches fourfold, reducing the data footprint for light clients such as Helios. Swapping the hash function for BLAKE3 or a variant of Poseidon would further improve proof efficiency by three to 100 times. Buterin acknowledged that Poseidon requires additional security auditing.<\/p>\n<p>The co-founder\u2019s proposal is an evolution of earlier plans to deploy Verkle trees. Interest in binary structures was revived in mid\u20112024 amid concerns that current cryptographic methods <a href=\"https:\/\/forklog.com\/en\/news\/ethereum-to-implement-quantum-threat-protection\">could prove vulnerable<\/a> to quantum computers.<\/p>\n<h2 class=\"wp-block-heading\">A new virtual machine<\/h2>\n<p>Buterin again raised <a href=\"https:\/\/forklog.com\/en\/news\/vitalik-buterin-proposes-shifting-ethereums-architecture-to-risc-v\">the question of replacing the EVM with RISC\u2011V<\/a>\u2014an open instruction-set architecture that most ZK provers already use. He outlined a three\u2011stage plan:<\/p>\n<ol class=\"wp-block-list\">\n<li>Introduce RISC\u2011V solely for precompiled contracts.<\/li>\n<li>Allow users to deploy contracts on the new architecture.<\/li>\n<li>Fully deprecate the old EVM, turning it into a smart contract running atop the new virtual machine.<\/li>\n<\/ol>\n<blockquote class=\"wp-block-quote is-layout-flow wp-block-quote-is-layout-flow\">\n<p><em>&#8220;The point of Ethereum is its universality. If the EVM is not good enough to achieve that, we should address the problem directly and build a better virtual machine,&#8221; the programmer wrote.<\/em><\/p>\n<\/blockquote>\n<p>The initiative has met resistance. In November 2025, researchers at Offchain Labs (the developers of <a href=\"https:\/\/forklog.com\/en\/news\/what-is-arbitrum\">Arbitrum<\/a>) <a href=\"https:\/\/forklog.com\/en\/news\/offchain-labs-criticizes-buterins-proposal-to-transition-evm-to-risc-v\">argued<\/a> that WebAssembly (WASM) is a better fit for long\u2011term development.<\/p>\n<p>Their key point: RISC\u2011V excels at <a href=\"https:\/\/forklog.com\/en\/news\/what-is-a-zero-knowledge-proof\">ZK proofs<\/a>, but an &#8220;interface architecture&#8221; and an &#8220;architecture for proofs&#8221; do not have to coincide.<\/p>\n<h2 class=\"wp-block-heading\">AI<\/h2>\n<p>Buterin also suggested that delivering on <a href=\"https:\/\/forklog.com\/en\/news\/ethereum-developers-plan-seven-hard-forks-by-2029\">the Ethereum roadmap<\/a> could accelerate thanks to <span data-descr=\"a software\u2011development method where the programmer writes little code by hand and delegates code generation, fixing and testing to neural networks by describing the desired outcome in natural language\" class=\"old_tooltip\">vibe\u2011coding<\/span>.<\/p>\n<blockquote class=\"twitter-tweet\">\n<p lang=\"en\" dir=\"ltr\">This is quite an impressive experiment. Vibe-coding the entire 2030 roadmap within weeks.<\/p>\n<p>Obviously such a thing built in two weeks without even having the EIPs has massive caveats: almost certainly lots of critical bugs, and probably in some cases &#8220;stub&#8221; versions of a thing\u2026 <a href=\"https:\/\/t.co\/ZlTg0r2hvI\">https:\/\/t.co\/ZlTg0r2hvI<\/a><\/p>\n<p>\u2014 vitalik.eth (@VitalikButerin) <a href=\"https:\/\/twitter.com\/VitalikButerin\/status\/2027781104288317837?ref_src=twsrc%5Etfw\">February 28, 2026<\/a><\/p><\/blockquote>\n<p> <script async src=\"https:\/\/platform.twitter.com\/widgets.js\" charset=\"utf-8\"><\/script><\/p>\n<p>In his words, AI &#8220;significantly accelerates writing code&#8221;. The community should be open to the possibility that the project\u2019s plans will be delivered &#8220;much faster and with a higher level of security than expected&#8221;.<\/p>\n<p>He was commenting on the results of an experiment in which a developer built a reference implementation of the roadmap in just a few weeks.<\/p>\n<blockquote class=\"twitter-tweet\">\n<p lang=\"en\" dir=\"ltr\">Two weeks ago I made a bet with <a href=\"https:\/\/twitter.com\/VitalikButerin?ref_src=twsrc%5Etfw\">@VitalikButerin<\/a> that one person could agentic-code an <a href=\"https:\/\/twitter.com\/ethereum?ref_src=twsrc%5Etfw\">@ethereum<\/a> client targeting 2030+ roadmap. So I built ETH2030 (<a href=\"https:\/\/t.co\/2k83PyUP4z\">https:\/\/t.co\/2k83PyUP4z<\/a> | <a href=\"https:\/\/t.co\/P0A6aHDZBX\">https:\/\/t.co\/P0A6aHDZBX<\/a>).<\/p>\n<p>702K lines of Go. 65 roadmap items. Syncs with mainnet. Here&#8217;s what I found. <a href=\"https:\/\/t.co\/i6gkqLc82L\">https:\/\/t.co\/i6gkqLc82L<\/a><\/p>\n<p>\u2014 YQ (@yq_acc) <a href=\"https:\/\/twitter.com\/yq_acc\/status\/2026252944639934778?ref_src=twsrc%5Etfw\">February 24, 2026<\/a><\/p><\/blockquote>\n<p> <script async src=\"https:\/\/platform.twitter.com\/widgets.js\" charset=\"utf-8\"><\/script><\/p>\n<p>However, Buterin pointed to &#8220;large caveats&#8221;: the high speed of coding &#8220;almost guarantees the presence of critical bugs&#8221;, and in some cases the AI may have produced only &#8220;stubs&#8221; instead of working versions.<\/p>\n<blockquote class=\"wp-block-quote is-layout-flow wp-block-quote-is-layout-flow\">\n<p><em>&#8220;But half a year ago even such results were unattainable, so for now the direction of the trend matters more,&#8221; he added.<\/em><\/p>\n<\/blockquote>\n<p>In his view, the optimal approach is to use neural networks to strengthen security, not only to speed things up. Possible measures include generating additional test scenarios, applying formal verification and creating alternative implementations of the same components.<\/p>\n<p>In February, Buterin <a href=\"https:\/\/forklog.com\/en\/news\/buterin-envisions-future-alliance-between-ai-and-ethereum\">described<\/a> a possible process for merging Ethereum with artificial intelligence.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Vitalik Buterin set out two execution-layer changes for Ethereum: a binary state tree and a long-term EVM replacement.<\/p>\n","protected":false},"author":1,"featured_media":94823,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"select":"1","news_style_id":"1","cryptorium_level":"","_short_excerpt_text":"Buterin proposes a binary state tree and a long-term EVM replacement for Ethereum.","creation_source":"","_metatest_mainpost_news_update":false,"footnotes":""},"categories":[3],"tags":[46,433,200],"class_list":["post-94822","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-news-and-analysis","tag-ethereum","tag-evm","tag-vitalik-buterin"],"aioseo_notices":[],"amp_enabled":true,"views":"186","promo_type":"1","layout_type":"1","short_excerpt":"Buterin proposes a binary state tree and a long-term EVM replacement for Ethereum.","is_update":"","_links":{"self":[{"href":"https:\/\/forklog.com\/en\/wp-json\/wp\/v2\/posts\/94822","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/forklog.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/forklog.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/forklog.com\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/forklog.com\/en\/wp-json\/wp\/v2\/comments?post=94822"}],"version-history":[{"count":1,"href":"https:\/\/forklog.com\/en\/wp-json\/wp\/v2\/posts\/94822\/revisions"}],"predecessor-version":[{"id":94824,"href":"https:\/\/forklog.com\/en\/wp-json\/wp\/v2\/posts\/94822\/revisions\/94824"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/forklog.com\/en\/wp-json\/wp\/v2\/media\/94823"}],"wp:attachment":[{"href":"https:\/\/forklog.com\/en\/wp-json\/wp\/v2\/media?parent=94822"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/forklog.com\/en\/wp-json\/wp\/v2\/categories?post=94822"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/forklog.com\/en\/wp-json\/wp\/v2\/tags?post=94822"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}