Architectural Overview

Eclipse is designed as an optimistic layer 2 blockchain on top of Ethereum, using Celestia to post data, guaranteeing data availability. Let’s break this down further:

• Layer 2 rollup: Eclipse employs a permissioned set of sequencers responsible for ordering and executing transactions off-chain. Periodically, each sequencer bundle produces a state commitment, a succinct cryptographic summary of the Layer 2 chain’s current state, which is posted to an Ethereum smart contract. This approach inherits Ethereum’s integrity guarantees (e.g., finality, censorship resistance) without requiring a large on-chain validator set. For instance, censorship resistance can be achieved through the use of the forced queue mechanism.

• Optimistic fraud proofs: Eclipse doesn't provide algorithmic proof of transaction execution when posting data to Ethereum. Instead, it relies on fraud proofs (pending deployment) for network safety. This means that eventually, anyone would be able to challenge the validity of state transition statements posted on the Eclipse smart contract by proving they're invalid. The approach we have chosen for fraud proofs relies on zero-knowledge technology for responding to potential fraud proof challenges. This design keeps operational costs low, while inheriting strong security guarantees of Ethereum to prevent adversarial behavior.

• Data Availability: As a rollup, Eclipse is required to post all the data required to re-execute blocks to a platform that guarantees the availability of this data even under different censorship scenarios. Such platforms are called data availability (DA) layers. While some rollups use Ethereum’s EIP-4844 blog storage for data availability, Eclipse uses a DA layer called Celestia, because of higher throughput and more favorable pricing. Celestia’s data‐availability network is optimized for batched rollup data at scales that exceed Ethereum’s EIP-4844 blob storage limits. The reader is referred to our blog post on choosing a DA layer.