Ethereum's L1 roadmap drops the Poseidon hash for SHA-2 or BLAKE2s; production leanVM targeted for 2027

Ethereum Foundation/Lean Ethereum

Ethereum Foundation researcher Justin Drake said Ethereum's base layer will no longer plan around Poseidon, a hash function designed to be cheap inside zero-knowledge proofs, and will use long-established hashes such as SHA-2 or BLAKE2s instead. Newer proof systems have become fast enough at proving ordinary hashes that the special-purpose one is no longer worth its weaker track record. It closes what Drake called an "8-year, 8-figure" research detour on the road to post-quantum Ethereum.

• Why it changed: binary-field SNARKs now prove about 1 million standard hash calls per second on a laptop. Drake: "the key was not SNARK-friendly hashes, but hash-friendly SNARKs."
• Older, heavily studied hashes carry less risk of a hidden break; recent AI-assisted cryptanalysis found weaknesses in newer post-quantum schemes such as HAWK and SQIsign.
• Timeline: a production-grade leanVM (the proving machine for the redesign) in 2027, then deployment across Ethereum's consensus, data and execution layers in 2028.
• Scope: only Ethereum's own base-layer roadmap. Rollups and other projects that already use Poseidon are not asked to replace it.