Rollups

What are Rollups?

Rollups have emerged as the dominant approach to blockchain scaling, fundamentally reshaping how decentralized networks handle transaction throughput. At their core, rollups are Layer 2 solutions that execute transactions outside the main blockchain while posting transaction data or validity proofs back to the base layer. This architecture enables dramatically higher throughput and lower costs while preserving the security guarantees that make the underlying blockchain valuable.

The key insight behind rollups is that consensus and data availability are more expensive on-chain operations than computation. By moving execution off-chain while keeping data on-chain, rollups achieve the best of both worlds. Thousands of transactions can be bundled together and represented by a single commitment on the base layer, spreading the fixed costs across many operations and reducing per-transaction fees by orders of magnitude.

What distinguishes rollups from other scaling approaches is their security inheritance from the base layer. Unlike sidechains that rely on their own validator sets, rollups ensure that users can always recover their funds through the main chain even if all rollup operators disappear or act maliciously. This strong security guarantee, combined with practical throughput improvements, has made rollups the centerpiece of Ethereum’s scaling roadmap and inspired similar efforts across other blockchain ecosystems.

How Rollups Work

The fundamental operation of a rollup begins when users deposit assets into a smart contract on the base layer. This bridge contract locks funds on Layer 1 and enables equivalent assets to be minted on the rollup, where users can then transact freely. The rollup maintains its own state, tracking account balances, smart contract storage, and all the information necessary to process transactions independently of the main chain.

A component called the sequencer collects incoming transactions, orders them, and executes them to produce new state. After processing a batch of transactions, the sequencer computes a new state root, a cryptographic commitment that summarizes the entire rollup state. This state root, along with compressed transaction data, gets posted to the base layer. The transaction data posting is crucial because it ensures anyone can independently reconstruct the rollup state and verify its correctness.

The security of this architecture depends on data availability. Because all transaction data is published on-chain, any observer can detect if a sequencer posts an invalid state root. The specific mechanism for challenging or proving correctness differs between rollup types, but the underlying guarantee remains the same: the rollup inherits the security of its base layer because verification happens on-chain where the full economic weight of the network applies.

Optimistic vs ZK Rollups

The two primary families of rollups differ fundamentally in how they ensure transaction validity. Optimistic rollups assume transactions are valid by default and rely on a challenge mechanism called fraud proofs. After a batch is posted, a window of typically seven days allows anyone to submit proof that the batch contains incorrect state transitions. If no valid challenge appears, the batch becomes final. This approach requires minimal on-chain computation during normal operation, as the expensive fraud proof process only triggers when someone detects an error.

ZK rollups take the opposite approach, generating cryptographic validity proofs that mathematically demonstrate correct execution before posting results to the chain. Rather than waiting for challenges, ZK rollups prove validity upfront using zero-knowledge cryptography. The base layer verifies these proofs, which is computationally cheap despite the complex mathematics involved, and immediately considers the batch final. This enables much faster withdrawals and stronger security guarantees at the cost of more complex and computationally intensive proof generation.

The trade-offs between these approaches continue to evolve as technology advances. Optimistic rollups achieved EVM compatibility earlier and currently handle more transaction volume, benefiting from simpler implementation and lower operational costs. ZK rollups offer superior finality speed and potentially stronger security properties, but generating validity proofs for arbitrary EVM execution remains technically challenging. Many observers expect ZK technology to eventually dominate as proof systems mature, though optimistic rollups may retain advantages for specific use cases.

Rollup Economics

The economic model of rollups revolves around the sequencer, which earns revenue from transaction fees while incurring costs for posting data to the base layer. Sequencers collect fees from users similar to how miners or validators earn on Layer 1, but their primary expense is the cost of publishing transaction data on-chain. This creates a natural incentive to batch many transactions together, spreading the fixed posting cost across more fee-paying operations.

Data costs represent the largest operational expense for most rollups and directly influence user transaction fees. The base layer charges for every byte of data published, creating a floor below which rollup fees cannot sustainably drop. Recent developments in data availability, including Ethereum’s EIP-4844 proto-danksharding upgrade, have introduced cheaper dedicated data space for rollups, significantly reducing costs. Future improvements like full danksharding promise to expand data availability capacity further.

MEV (Maximal Extractable Value) considerations add another dimension to rollup economics. Sequencers who order transactions can potentially extract value through frontrunning, sandwiching, or other ordering-dependent strategies. Most current rollups operate with centralized sequencers, raising concerns about MEV extraction and censorship. Various proposals aim to address these concerns through sequencer decentralization, fair ordering protocols, or MEV redistribution mechanisms that return extracted value to users or the protocol.

Major Rollup Implementations

Arbitrum One has grown to become the largest rollup by total value locked and daily transaction volume. Developed by Offchain Labs, it implements a sophisticated interactive fraud proof system and has attracted a comprehensive ecosystem of DeFi protocols, gaming applications, and infrastructure projects. Arbitrum’s success demonstrates that optimistic rollups can achieve meaningful scale while maintaining security, processing millions of transactions at costs far below Ethereum mainnet.

Optimism pioneered much of the optimistic rollup architecture and continues driving ecosystem-wide innovation. Beyond its own network, the Optimism team developed the OP Stack, an open-source modular framework for building optimistic rollups. This framework has been adopted by Base, operated by Coinbase, along with numerous other projects, creating a family of interoperable Layer 2 networks called the Superchain. The shared codebase and standards enable easier cross-rollup communication and reduce fragmentation.

On the ZK side, zkSync Era and StarkNet represent leading implementations with different technical approaches. zkSync, built by Matter Labs, uses a custom virtual machine and LLVM-based compiler to achieve Solidity compatibility while optimizing for proof efficiency. StarkNet, developed by StarkWare, employs STARK proofs with no trusted setup and a purpose-built language called Cairo. Polygon zkEVM and Scroll focus more directly on bytecode-level EVM equivalence, allowing existing contracts to deploy without modification. Each approach makes different trade-offs around compatibility, performance, and security properties.

Future of Rollups

Based rollups represent an emerging design that replaces centralized sequencers with the base layer’s own proposers. Rather than relying on a dedicated sequencer to order transactions, based rollups allow Layer 1 validators to directly include rollup blocks. This approach inherits the base layer’s decentralization and censorship resistance while eliminating the need for separate sequencer infrastructure. The trade-off involves potentially higher latency and less sophisticated transaction ordering, but the security and decentralization benefits may outweigh these costs for many applications.

Shared sequencing networks offer another path toward sequencer decentralization while enabling powerful cross-rollup capabilities. Multiple rollups could delegate ordering to a common decentralized sequencer set, which would provide consistent ordering across all participating chains. This shared ordering enables atomic transactions spanning multiple rollups, where operations on different chains either all succeed or all fail together. Such composability would dramatically reduce the fragmentation that currently makes moving between rollups cumbersome.

Cross-rollup composability remains one of the most important challenges for the maturing rollup ecosystem. Users currently face friction when assets or applications they need exist on different rollups, requiring bridge transactions and fragmented liquidity. Solutions under development include standardized message passing protocols, unified bridge infrastructure, and user interfaces that abstract away the underlying rollup complexity. The vision of seamless interaction across a constellation of specialized rollups, sometimes called rollup abstraction, would preserve the benefits of modularity while delivering a unified user experience.

Conclusion

Rollups have transformed from experimental scaling concepts to essential blockchain infrastructure, processing the majority of Layer 2 transaction volume and enabling applications that would be economically unviable on congested base layers. By executing transactions off-chain while posting data on-chain, they achieve the throughput necessary for mainstream adoption without sacrificing the security properties that make blockchains valuable.

Understanding rollups is increasingly important for anyone building or using blockchain applications. The choice between different rollup implementations involves trade-offs around finality speed, EVM compatibility, decentralization, and ecosystem maturity that vary by use case. As the technology continues evolving through improved proof systems, decentralized sequencing, and cross-rollup interoperability, rollups will likely remain the primary execution environment for blockchain activity, with the base layer increasingly serving as a settlement and data availability foundation for a diverse ecosystem of specialized rollup networks.