Ethereum Foundation researcher Justin Drake announced that Ethereum's Layer 1 cryptography roadmap is abandoning Poseidon, the zero-knowledge-friendly hash function the protocol has relied on since 2019, in favor of traditional, extensively studied functions such as SHA or BLAKE.
Drake described the shift as the conclusion of an “epic 8-year, 8-figure rabbit hole in post-quantum cryptography,” framing it as a milestone that unlocks stronger long-term security guarantees for what he called “lean Ethereum” — the protocol's ongoing effort to simplify its core architecture.
Why Poseidon Is Being Retired
Poseidon was designed specifically to be cheap to compute inside zero-knowledge proof systems known as SNARKs, a property that made it attractive for Ethereum's scaling roadmap when general-purpose hash functions were still too expensive to use efficiently inside those proofs. According to Drake's own announcement, breakthroughs in SNARK design mean that off-the-shelf functions like SHA2 and BLAKE2s can now match Poseidon's performance inside a SNARK, removing the technical justification for keeping a specialized, less battle-tested primitive in the protocol's critical path.
That distinction matters because SHA and BLAKE have each undergone extensive cryptanalysis outside of zero-knowledge contexts for years, while Poseidon, as a newer and more narrowly used function, carried comparatively less independent scrutiny — a gap that becomes more consequential the longer a cryptographic primitive is expected to secure a network worth hundreds of billions of dollars.
Related: FG Nexus Dumps Its Entire ETH Treasury Less Than a Year In
Part of a Broader Post-Quantum Push
The move ties directly into Ethereum's post-quantum security program, which is preparing the protocol for a future in which sufficiently powerful quantum computers could break the elliptic-curve cryptography Ethereum and most other blockchains currently rely on for signatures. Relying on well-studied traditional hash functions reduces the number of unproven cryptographic assumptions baked into that transition, even as the specific post-quantum signature schemes themselves remain a separate, ongoing area of research.
Under the current roadmap, a production-grade “leanVM” is targeted for 2027, with consensus-layer, data-layer and execution-layer deployments following in 2028 — a timeline that gives the Ethereum research community roughly two years to finalize the switch before it reaches mainnet-critical components.
