Cascade: The First Interchain SVM
Cascade features an innovative new rollup architecture design.
There are two core ideas that form the basis of Cascade:
- 1.Optimistic rollups
- 2.Sovereign rollups
In this article, we'll explain how optimistic rollups and sovereign rollups on a high level.
A typical blockchain (or "rollup") engages in the following functions:
- Execution: computing the state of the blockchain after ingesting transactions
- Settlement (for rollups): verifying the correctness of execution and bridging
- Consensus & data availability: establishing the ordering of transactions
At a high level, rollups execute transactions and separately must "settle" that execution to verify that the execution was correct.
There are two types of settlement: optimistic or zero-knowledge. An optimistic rollup periodically posts the result of transactions to some settlement layer along with a bounty. If a challenger correctly asserts that the execution was invalid, the challenger receives some (or all) of the bounty. The settlement layer must have some method of determining whether a challenge is correct or not.
A "smart contract rollup" like Optimism or Arbitrum settles to some set of smart contracts on a Layer 1 blockchain like Ethereum. By contrast, "sovereign rollups" don't require such smart contracts to perform settlement. Instead, sovereign rollups like the Eclipse settlement layer circulate fraud proofs directly to light clients. If the majority of Eclipse becomes dishonest, these circulated fraud proofs ensure that an honest minority can still maintain the integrity of the chain.
Cascade block production is primarily managed by a single party, called the "sequencer," which helps the network by providing the following services:
- Providing transaction confirmations and state updates
- Constructing and executing rollup blocks
- Posting rollup blocks to the settlement layer, where they are passed to the consensus layer
For the moment, the Eclipse team runs the primary block producer. Over time, there are plans to further decentralize the sequencer role. In order to further increase scalability parameters, there are plans to integrate Celestia in the near future to serve as a modular data availability layer.