Single-slot finality (SSF) in Ethereum is an advanced concept designed to dramatically reduce the time it takes for a block to achieve finality. In the current Ethereum framework, it takes approximately 15 minutes for a block to be finalized, which means it becomes a permanent part of the blockchain and cannot be altered or removed without significant cost. SSF, however, proposes a system where blocks can be both proposed and finalized within the same time slot, offering near-instantaneous finality.
Ethereum Finality
Finality in Ethereum's context refers to the assurance that once a block is added to the blockchain, it is irreversible. This is achieved through Ethereum's Proof-of-Stake consensus mechanism, where validators stake ETH as collateral. If they act maliciously or try to manipulate the blockchain, they risk losing their staked ETH. This crypto-economic model deters rational actors from attempting to alter the blockchain's order or content.
In the post-merge environment, Ethereum utilizes a new consensus mechanism named Gasper. Gasper combines Casper FFG (how to determine finality) and LMD GHOST (determining fork-choice rules) to achieve consensus. Similar to its old Nakamoto-style consensus in which liveness was prioritized over safety, Gasper is also a liveliness-preferring consensus protocol. Consensus algorithms that prioritize safety (such as Tendermint) fail if they do not receive the required amount of votes (2/3 of the validator set). This leads to the chain halting (the opposite of "liveness"). Chains that prioritize liveness will continue to generate blocks based on their "heaviest/longest chain" rule set no matter what, even if bad actors have corrupted the majority of the hash power/votes.
Casper the Friendly Finality Gadget is the name of Beacon Chain's PoS-based consensus protocol. Casper is a partial consensus algorithm that combines PoS and BFT consensus models. Casper's core design was modeled after the PBFT consensus method but, similar to Nakamoto Consensus, still identifies the "true" chain as the one with the greatest attestations. Casper FFG decides on which blocks become part of the chain.
Casper reaches finality through periodic checkpoints after a sufficient number of votes from its 470,000+ validators. Checkpoints can be subdivided into slots and epochs. A slot is when a block is added to the Beacon chain, and in PoS, it occurs predictably every 12 seconds. A slot is divided into three equal parts, with each part spanning four seconds. At the commencement of each slot, a validator is selected at random to propose a block, hereby referred to as the 'proposer'. Concurrently, other validators are entrusted with the responsibility of attesting or casting votes for the block that they perceive as the chain head, guided by their local view and the fork-choice rule.
The attestation deadline at t=4 is a pivotal moment in the slot. Should a validator tasked with attestation not receive a block by the deadline, the validator would vote for the previously accepted head of the chain as per the fork-choice rule. The propagation speed of a block, therefore, plays a significant role in the number of attestations it receives. Simply put, the sooner a block is proposed, the more time it has to circulate, thereby accruing more attestations.Optimal network health is achieved when a block is published at t=0, as per the Ethereum specification. However, the value of a block increases steadily with time, thus incentivizing proposers to delay their block's publication, allowing for the accumulation of more Maximum Extractable Value (MEV).
A different validator is assigned to propose a new block in each slot ahead of time. 32 slots (6.4 min) are grouped into an epoch. Each slot also has a group of validators (minimum of 128) called a committee assigned to it. Committees verify and attest to the validity of each proposed block. After successfully verifying the block, the committee members broadcast a cryptographic attestation of the block to the other validators and network nodes. Based on the validator attestations, the fork-choice rule LMD GHOST decides the current head of the chain. Generally, finality is attained with the necessary votes after two epochs.
Epochs are what is considered in the periodic checkpoints and for chain finalization. An epoch becomes justified if> 66% of the validators attest during that epoch. If a second epoch with > 66% follows the first, it will finalize that epoch and make the transactions irreversible.
Source: Consensys
Single-slot Finality
The push for SSF arises from the need to address limitations inherent in the current finality timeframe. A 15-minute delay is often too long for many users and applications, especially those that require high transaction throughput, like exchanges. This delay also opens a window for potential short reorganizations (reorgs) of the blockchain, which could be exploited for nefarious purposes like censorship or extracting maximum extractable value (MEV). Additionally, the existing multi-stage block upgrade mechanism is complex and has been susceptible to security vulnerabilities, necessitating multiple patches.
Implementing SSF would entail optimizing Ethereum's consensus mechanism and block processing capabilities to allow for the rapid proposal and finalization of blocks. This would enhance transaction throughput, provide near-instant finality, improve security against various attacks, and simplify Ethereum's upgrade process.
The Path to Achieving SSF:
Advances in Signature Aggregation: The development of signature aggregation schemes, particularly Boneh-Lynn-Shacham (BLS), offers scalability beyond initial expectations. Improved client capabilities in processing and verifying these signatures have made it feasible to handle attestations from a large number of validators within a single slot.
Signature Verification and Processing Capacity: For example, with a million validators each voting twice in a 16-second slot, nodes would need to process signatures at a rate of 125,000 aggregations per second. With current technology, a typical computer can verify a signature in about 500 nanoseconds, making this rate achievable.
Super-Committees for Efficiency: To manage this, the concept of 'super-committees' has been proposed, where a randomly selected subset of validators (e.g., 125,000 per slot) would have the authority to finalize a block. This approach balances the attack cost with scalability. An attacker would need to control a significant portion of the staked ETH in a supercommittee to successfully compromise a block. Adjusting the size of the validator set or the supercommittee can modify the attack's cost.
- Challenges in Signature Aggregation: While signature verification is manageable, the aggregation of these signatures poses a significant challenge. Solutions may involve increasing validators in subnets, expanding the number of subnets, or introducing additional aggregation layers. Specialized aggregators could also be employed, akin to the proposer-builder separation (PBS) and Danksharding frameworks, where specific tasks are outsourced to experts.
The journey to single-slot finality in Ethereum involves ongoing research and discussion within the community. The objective is to optimize signature aggregation processes, balancing scalability, security, and the cost of potential attacks. These efforts are crucial to enhancing Ethereum's consensus mechanism, aiming to provide faster finality while maintaining the network's robustness and integrity.
