Unraveling Ethereum: A Comprehensive Guide to L1 and L2 Scaling Solutions

Rollup is a network that works parallel with Ethereum but records transactions into the main chain. Validity checks for rollup transactions are performed within Ethereum.
There are two types of rollups, depending on the method of transaction checks: optimistic and zk-rollups.

Optimistic rollups
Optimistic rollups form batches from the transactions and then record them into the main network in a compressed way. This allows the recording to have a lower gas fee since the transactions aren’t released directly in the Ethereum network and because of the format of the recordings. So the whole process becomes way more affordable for an average user. Optimistic rollups allow the L2 transaction data recorded into the network to be considered valid by default while leaving some time to dispute said data. There is a risk that if an invalid transaction isn’t disputed, it might be recorded to L1 as a valid one anyway.

ZK-rollups
ZK-rollups are also networks that function simultaneously with Ethereum. However, the validation process is different for them. Batches are formed from transactions, but the data that gets recorded to L1 is a summary of these batches as well as cryptographic proof that the transactions are valid. At this time, ZK-rollups are quite possibly the most promising scaling solution for Ethereum. It’s in many ways because, unlike their Optimistic counterparts, they require transaction data to be cryptographically validated. There’s also a smaller waiting period since there is no need to wait for any disputes to show up or expect the transaction to be canceled. At the same time, the mathematical complexity of ZK-proofs places its own restrictions: validation of general-purpose EVM calculations is a complicated task in and of its own. Thankfully, a lot of developers are focusing their work on solving this issue.

State channel is a technology that allows a group of users to exchange transactions among themselves while recording only two of them – the first and the last one, into the main network.
Here’s how it works: a multisig smart contract is deployed into Ethereum. This contract checks that the transactions have been signed by all the required participants. Then the members of the state channel deposit their funds into this multisig contract. After that, they perform all the required interactions off-chain and sign the result. As the final step, the smart contract distributes the funds between the participants according to the recorded result.

Plasma chains are something of a middle ground between rollups, where a full validation of the submitted transactions is run within the L1, and side chains without such validation. The idea of plasma chains lies in the premise that not all transactions require validation by all nodes on Ethereum.

Occasionally, a plasma chain records a cryptographic proof of the network condition along with the result of its work. The data itself isn’t recorded and the proof is compact. In other words, the validity of the transactions isn’t checked but if the commitment is already recorded on Ethereum, a plasma chain can’t change the transaction history retroactively.

In terms of architecture, Validiums are a lot like ZK-rollups with one important difference: transaction data for validation is stored off-chain. This allows to provide high throughput with lower fees. However, that is achieved at the cost of security. ZK-rollups deliver higher security. Without data stored in the L1, a Validium operator can prevent a user from using their funds.