Ethereum researchers have submitted a draft protocol proposal that would separate the votes used to choose the chain head from those used to finalize blocks. The Core proposal, labeled EIP-8453 and titled Decoupled Consensus, was opened on October 21 by Ben Edgington with Francesco D'Amato, Roberto Saltini, Yann Vonlanthen and Mikhail Kalinin. Its initial target is at least a fourfold reduction in time to finality. The change is not active: the pull request remains open and unmerged, and its authors say important parameters and security work are still incomplete.

Two voting paths instead of one

Ethereum's current Gasper design combines LMD-GHOST fork choice with Casper FFG finality. A validator attestation helps identify the chain that nodes should build on while also contributing to the slower process that makes earlier blocks final. EIP-8453 would turn those functions into separately propagated and processed messages. A smaller available-chain committee, described in the draft as approximately 512 validators per slot, would cast head votes. The full validator set would provide finality attestations on a separate round schedule.

Related reporting: Ethereum Sets October 26 Hoodi Test for Glamsterdam Upgrade

Why finality time matters

Finality is the point at which the protocol treats a block as irreversible unless a large amount of staked ether is destroyed through penalties. CoinDesk reported that Ethereum transactions currently take about 16 minutes to reach expected finality. A fourfold improvement would bring that interval to roughly four minutes, although the proposal does not yet fix the parameters needed to deliver that result. Faster finality could reduce the period during which exchanges, bridges and payment services must hold capital against transfers that have arrived but are not yet fully settled.

The security case

The authors also want to limit the influence of old fork-choice votes. The draft says LMD-GHOST retains validators' latest votes without a fixed age limit, creating an opportunity for withheld votes to be released strategically. Its proposed available chain would emphasize recent votes and use expiring locks to protect confirmed blocks when participation falls. Finality would advance when included attestations reach a quorum of at least two-thirds of active stake. Available-chain confirmations would still depend on network timing and participation and would not carry the same guarantee as finality.

Trade-offs remain unresolved

Sampling a smaller head-vote committee could reduce the signature workload inside each slot, but the proposal acknowledges new risks. A smaller group can become a denial-of-service target, and sampling introduces uncertainty about whether the committee represents the wider validator set. The selection method and committee size are not fixed. The draft also leaves vote encoding, aggregation, timing, rewards, penalties, inactivity handling and fork-transition rules to future work. A deployment would require a consensus-layer hard fork and updates to consensus and validator clients.

A roadmap step, not a launch

The proposal is consistent with the Ethereum Foundation's published fast-finality direction, which calls for decoupling finality from block production through an available chain and a separate finality gadget. EIP-8453 also aims to make the consensus layer more flexible for a future post-quantum transition, but it does not choose a post-quantum signature scheme. The pull request's immediate purpose is to establish and review the standards-track proposal. Before any mainnet decision, developers would need a complete specification, executable test vectors, formal safety and liveness analysis, client implementations and agreement on an upgrade schedule.

Sources

AI-generated editorial image; not a photograph of the reported event. Prepared with AI assistance and source verification.