CryptoSlate
Ethereum researchers are racing to close a 52-bit gap in zkEVM cryptographic security before a December 2026 review. The better.codes contest's live leaderboard, captured at 15:44:47 UTC on Aug. 21, showed a 63.99-bit soundness certificate and a 116.13-bit attack certificate for koalaIRS12, leaving 52.14 bits unresolved between the two promoted results after nine submissions from seven solvers. KoalaIRS12 is a fixed parameter profile for an interleaved Reed-Solomon reduction used in proof-system research, not the full zkEVM itself. The open interval sits well below Ethereum's 128-bit provable-security target for the broader system.
Why it matters
The certificates are machine-checked bounds inside a pinned Lean environment, where soundness submissions lift the lower certificate and attack submissions pull the upper certificate down. Each promotion moves a reproducible boundary, but the framework explicitly forbids reading either certificate as a full-system soundness claim. A production-grade zkEVM would need additional layers: completeness, the assumptions baked into formal definitions, fidelity between specifications and implementations, and proof composition across separately analyzed components.
A May review of an SP1 formal-verification effort showed why those layers matter. Specifications and theorem statements are themselves code, components need versioned pinning, and individual results require broader reasoning before they can support a system claim. An academic paper by Gal Arnon, Dan Boneh, and Giacomo Fenzi separately flagged list decoding, Reed-Solomon proximity gaps, and correlated-agreement questions as open problems for succinct proof systems.
Market impact
The December 2025 zkEVM security roadmap called for 128-bit provable security, proof sizes at or below 300 KiB, and a formal soundness argument for the recursion architecture. A February update pushed the M3 deadline to early December 2026, with the architecture-security argument now anchored to a Dec. 1 deliverable. For koalaIRS12 alone, a lower certificate reaching 128 bits would settle just the soundness side of this benchmark at its fixed parameter point.
Frequently asked questions
-
What is the 52-bit gap in Ethereum's zkEVM?
It is the unresolved interval between a 63.99-bit soundness certificate and a 116.13-bit attack certificate for the koalaIRS12 benchmark, captured on the better.codes leaderboard at 15:44:47 UTC on Aug. 21 after nine submissions from seven solvers.
-
What is koalaIRS12?
KoalaIRS12 is a fixed parameter profile for an interleaved Reed-Solomon reduction used in zkEVM proof-system research. The better.codes contest scores certificates against this single profile, not the full Ethereum zkEVM.
-
What is the December 2026 M3 milestone?
M3 was originally set in the December 2025 zkEVM security roadmap and was pushed out in February. It calls for 128-bit provable security, proof sizes at or below 300 KiB, and a formal soundness argument for the recursion architecture by early December 2026.
-
Does the 52-bit gap affect Ethereum mainnet today?
No. A May Foundation update framed zkEVM proofs as non-consensus-critical option proofs. Mainnet attestation continues to rely on ordinary execution-client re-execution while zkEVM proofs supplement testing.
-
How does the better.codes contest close the interval?
Soundness submissions lift the lower certificate toward the 128-bit encoded target, while attack submissions pull the upper certificate down. Each accepted promotion is a machine-checked boundary change inside a pinned Lean environment.
CoinTelegraph
CoinDesk
Crypto Rank News
WuBlockchain