How rollups batch transactions

Ethereum can process only a limited number of transactions per block, so fees rise when demand is high. Rollups are the main way the ecosystem scales: they move execution off the main chain while still relying on Ethereum for security.

The basic idea

A rollup runs its own chain where users send transactions. A component called the sequencer orders those transactions and executes them. Instead of posting every transaction to Ethereum individually, the rollup compresses a large batch into a small amount of data and posts that to Ethereum along with the resulting state. Many users share the cost of one Ethereum transaction, which is why fees are lower.

Two ways to prove the batch is valid

• Optimistic rollups assume each batch is correct. Anyone can challenge it during a dispute window, commonly about seven days, by submitting a fraud proof.
• Zero-knowledge (ZK) rollups attach a cryptographic validity proof to each batch. Ethereum verifies the proof, so no challenge window is needed.

Where the data goes

For users to be able to rebuild the rollup's state and exit without the operator's help, the batch data must be available. Since Ethereum's 2024 upgrade, rollups can post this data in "blobs", a cheaper type of space that is kept for a limited time rather than forever.

The trade-offs

• Withdrawals: optimistic rollups usually need the dispute window to pass before funds move to Ethereum, unless you use a third-party bridge that fronts the liquidity.
• Complexity: ZK proofs are harder to build and generate, though they make finality faster.
• Centralisation: many rollups still run a single sequencer and have upgrade keys controlled by a small group. Check a rollup's current decentralisation stage before trusting it with large amounts.

Key takeaways

• Rollups execute off-chain and settle on Ethereum.
• Optimistic and ZK rollups differ in how they prove correctness.
• Lower fees come with trade-offs in withdrawal time and trust assumptions.