Ethereum researchers are preparing to move away from Poseidon, a hash function designed for zero-knowledge proofs, in favor of established algorithms including SHA-2 and BLAKE2 at the protocol’s base layer. Justin Drake, an Ethereum Foundation researcher, said in an Aug. 13 post on X that advances in binary-field SNARKs have removed a major reason for creating specialized “SNARK-friendly” hashes in the first place.
The change would redirect Ethereum’s zero-knowledge research toward proving systems that efficiently process conventional hashes, which have much longer public security records than Poseidon. It also supports the network’s planned post-quantum work, where hash functions could play a larger role in signature aggregation and verification.
Poseidon was introduced in 2019 and became common in zkRollups, zkVMs, and other SNARK-based applications. A SNARK, or succinct non-interactive argument of knowledge, lets a party prove that a computation was performed correctly without requiring every participant to repeat it. Poseidon gained traction because it required far fewer constraints than SHA-2 or Keccak when proofs were built over the large prime fields traditionally used by SNARKs.
Drake said the Ethereum ecosystem had spent eight years and tens of millions of dollars pursuing ZK-friendly hashing research. Yet the security trade-off has become harder to justify as Ethereum considers cryptographic systems intended to remain reliable through a potential transition to quantum-resistant technology.
SHA-2 and BLAKE2 have been examined extensively by academic and industry cryptographers, while newer specialized designs such as Poseidon have had less time to receive comparable scrutiny. Ethereum’s proposed direction would favor this longer-established cryptographic foundation without surrendering the performance gains that originally made Poseidon attractive.
Binary-field SNARKs change the hash calculation
The technical shift rests on binary-field SNARKs, a class of proof systems designed around arithmetic built from zeros and ones rather than the large prime-number fields used by many earlier SNARK designs.
Conventional hashes were historically expensive to prove because their internal logic relies heavily on Boolean operations such as XOR, bit shifts and rotations. Those operations map awkwardly into prime-field arithmetic circuits, creating a large proving cost for every hash invocation.
Binary fields allow those bit-level operations to translate more directly into proof constraints. Extension fields can then represent larger values while preserving the binary structure required for efficient Boolean computation. The result is a design path in which SNARK engineers can optimize their proof systems around widely deployed hashes instead of optimizing new hashes around existing proof systems.
Drake said a laptop can now verify roughly 1 million conventional-hash invocations per second inside newer binary-field constructions. He estimated that the resulting proving overhead is around 100 times native CPU Boolean computation, a substantial cost but one that he argued is increasingly practical for protocol-level cryptographic applications.
Research cited in the discussion includes Binius, a binary tower-field SNARK proposed in 2023 by Jim Posen and Benjamin Diamond. The work outlined a route for building proofs over binary fields, helping revive interest in proof systems that can handle standard hash functions more naturally.
A separate paper, Flock, uploaded to arXiv on July 29 by Benedikt Bünz, Ron Rothblum and William Wang, reported benchmark results on Apple’s M4 Max processor. According to the paper, Flock processed 82,000 BLAKE3 compressions per second per core, 42,000 SHA-256 compressions per second per core and 30,000 Keccak permutations per second per core. Its 10-core BLAKE3 result exceeded 660,000 compressions per second.
Drake also pointed to SNARK.fast, which he said recently reached 1.8 million BLAKE3 operations per second on an M3 Max. BLAKE3 is a newer member of the BLAKE family, while Ethereum’s stated base-layer discussion has focused more directly on SHA-2 and BLAKE2s.
Post-quantum plans place hashes at the center
The hash decision fits into Ethereum’s longer-term post-quantum roadmap. Drake’s proposed “strawmap” targets a production version of leanVM in 2027, followed by deployment across the consensus, execution and data-availability layers in 2028.
LeanVM is intended to help Ethereum compress large sets of post-quantum signatures into a small proof attached to each block. The approach centers on hash-based signatures, including a proposed scheme called leanXMSS.
Hash-based signatures are considered resistant to the type of quantum attack associated with Shor’s algorithm, which could break the elliptic-curve assumptions underlying Ethereum’s current BLS signature system. The trade-off is size: Ethereum’s post-quantum materials compare a leanXMSS signature of roughly 3,000 bytes with a 96-byte BLS signature. LeanVM’s target compression ratio is about 250-to-1, reducing the bandwidth burden of replacing compact BLS signatures with much larger quantum-resistant alternatives.
Ethereum has already started organizing development work around that possibility. The Ethereum Foundation launched the pq.ethereum.org resource hub in March 2026, while more than 10 client teams have participated in a weekly post-quantum interoperability devnet. The foundation has also offered a $1 million Poseidon Prize and a matching $1 million Proximity Prize for relevant cryptographic research.
Project Eleven’s The Quantum Threat to Blockchains — 2026 Report estimated that “Q-Day,” the point at which quantum computers could compromise ECDSA and RSA, may arrive between 2030 and 2033. Its estimate is uncertain by nature, but the timeline puts pressure on blockchains whose security upgrades require multiyear coordination across wallets, validators, client software and application infrastructure.
Other networks are testing their own routes
Ethereum is not alone in assessing post-quantum alternatives. The Solana Foundation published a post-quantum roadmap in April 2026. Solana validator teams Anza and Jump Crypto’s Firedancer independently selected Falcon, a post-quantum signature scheme standardized by the U.S. National Institute of Standards and Technology.
Solana’s plan outlined continued evaluation, possible use of quantum-resistant signatures for newly created wallets once the threat becomes material, and an eventual migration process for existing wallets. A separate system, Blueshift’s Solana Winternitz Vault, has operated for more than two years and was cited earlier in 2026 in a Google Quantum AI white paper.
StarkWare’s June 30 roadmap also starts with hash-function changes. The company said it plans to replace Pedersen hashing with BLAKE2 for state commitments, contract addresses and network configuration, while listing Falcon-512 among candidates for post-quantum consensus signatures.
Ethereum’s move away from Poseidon reflects a pragmatic recalculation: binary-field proof systems could allow protocol developers to retain zero-knowledge scalability while grounding critical hashing in algorithms with deeper cryptanalytic histories. The 2027 and 2028 targets remain ambitious, but the technical path now relies less on waiting for newer hash designs to mature and more on making established cryptography work efficiently inside modern proof systems.
Want deeper context on Ethereum’s evolution? Explore how it works, its upgrades, and trading impact in our Ethereum guide.
Disclaimer: The content on this page is provided for general informational purposes only and does not represent the views or financial advice of Toobit. We make no guarantees regarding the accuracy or completeness of this information and shall not be held liable for any errors, omissions, or outcomes resulting from its use. Investing in digital assets involves risk; users should independently evaluate their financial situation and the risks involved. For further details, please consult our Terms of Service and Risk Disclosure.
