What do you think of Ethereum’s “Rollup-centric” future?

Rollups solutions must prioritize reduced execution costs, cross-layer interoperability, and user privacy.

Key Takeaways

• To alleviate current scalability issues, Ethereum is moving to a "Rollup-centric roadmap" that combines the best on-chain and off-chain scaling solutions.

• Ethereum's on-chain scaling solution danksharding provides ample data space for the booming development of L2 (second layer) scaling solutions such as Optimistic Rollups and ZK-Rollups.

• Both Optimistic Rollups and ZK-Rollups have their shortcomings: most Optimistic Rollups have long withdrawal waiting periods between layers; and ZK-Rollups often require a lot of computation and cannot provide seamless composability.

• For a Rollup-centric future to remain viable, Rollup solutions must prioritize reduced execution costs, cross-layer interoperability, and user privacy.

Scalability is the method of increasing the speed and throughput of a network while minimizing (transaction) costs and without sacrificing the decentralization and security of a blockchain network. Ethereum is currently unable to scale. Over the past two years, as the number of users has increased, Ethereum's transaction costs have skyrocketed, making it nearly unaffordable for everyday investors to use the network.

There are two core options for Ethereum to scale: on-chain scaling and off-chain scaling. In conjunction with “The Merge”, the Ethereum team is introducing a new sharding design called “danksharding” to introduce an on-chain scaling solution. These changes are expected to occur sometime in the next few months.

Off-chain expansion comes from alternative protocols that sit on top of existing blockchains and do not require changes to the current Ethereum L1 protocol. L2 expansion solutions process all transaction verifications off-chain while obtaining security from the Ethereum mainnet. The most well-known L2 expansion solution is Rollups.

While preparing for the “merger,” Ethereum is moving toward a “Rollup-centric roadmap.” It plans to combine the best on-chain and off-chain scaling solutions.

01. A brief guide to Rollups

Rollups are the most prominent L2 scaling solution in the field. We can categorize Rollups based on their transaction verification and data storage methods. All Rollups aggregate large amounts of off-chain transaction data into batches and publish the results to the Ethereum mainnet when consensus is reached.

1) Optimistic Rollups

Optimistic Rollups networks, such as Arbitrum or Optimism, assume that all transactions are valid from the beginning. To ensure the security of these transactions, the Optimistic Rollups network provides a challenge period , in which network validators can question the legitimacy of a transaction through a fraud proof on the parent chain (such as Ethereum). By only executing proofs when fraud is suspected, Optimistic Rollups significantly increases throughput and reduces latency (transaction confirmation time). The challenge period is usually seven days.

Image source: ethereum.org

During the challenge period, users can use their assets within the L2 ecosystem, but cannot withdraw their assets back to the L1 chain. In response to the long withdrawal waiting period in Optimistic Rollups, some fast withdrawal solutions are emerging. For example, through a community-driven liquidity pool, Boba Network has shortened the withdrawal waiting period to a few minutes.

In addition, most Optimistic Rollups are EVM (Ethereum Virtual Machine) compatible , which means that anything you can do on L1, you can do in these Optimistic Rollups networks, and in a faster and cheaper way. EVM compatibility is the core reason why most Ethereum-based Rollups to date are Optimistic Rollups. For Ethereum developers, migrating smart contracts to these Optimistic Rollups solutions is a hassle-free process. In addition, Optimistic Rollups provide complete transparency because all transaction data is published to the Ethereum parent chain.

2) ZK-Rollups

Zero-knowledge (ZK)-Rollups networks, such as StarkNet, are similar to Optimistic Rollups in that they combine a large number of off-chain transactions and submit them back to the Ethereum mainnet in batches. However, instead of assuming that all transactions are valid until they are proven valid, ZK-Rollups use validity proofs to verify transactions immediately. These validity proofs and compressed data will be submitted to Ethereum L1 for on-chain verification as proxies for their corresponding original transaction packages.

Image credit: Simon Brown

Validity proofs are very complex and time-consuming, so ZK-Rollups have increased latency compared to Optimistic Rollups. Since generating cryptographic proofs (validity proofs) requires a lot of computation, the transaction sequencers in the ZK-Rollups network require high-specification hardware, making it difficult for everyday users to participate as their sequencers.

In addition, due to their inherent complexity, most ZK-Rollups are usually not EVM-compatible. As a result, it is difficult for Ethereum developers to migrate smart contracts to ZK-Rollups, which makes it more difficult to develop general applications for these ZK-Rollups without completely changing the smart contract framework. ZKSync and zkEVM have recently deployed EVM-compatible ZK-Rollup solutions on the Ethereum test network, bringing hope to a ZK-Rollup-centric future.

02. Ethereum’s “Rollup-centric” roadmap

At some point this year, Ethereum will undergo a shift in consensus mechanism from PoW (Proof of Work) to PoS (Proof of Stake), known as "The Merge." This change will reduce the Ethereum network's energy emissions, improve network security, and reduce centralization risks from miners. Ethereum's "Rollup-centric roadmap" will support any number of validators as long as they stake 32 ETH as an economic deposit.

Sharding is an on-chain horizontal scaling solution that increases the amount of activity a blockchain can efficiently handle. Essentially, the Ethereum blockchain will be split into smaller chains, so-called "shards," which will run in parallel and eventually be connected together through the Beacon Chain. Each shard chain will be run by a subset of nodes that will check data availability. This system ensures that anyone can run a network node using standard consumer-grade hardware while increasing the scalability of the network.

Instead of providing more space for transactions like traditional sharding, the latest danksharding scheme focuses on providing block space for data blobs. The Ethereum protocol does not attempt to interpret this data, but only verifies that these data blobs are fully available. For these data blobs to be considered fully available, no data must be lost and the Rollup state must be able to be rebuilt using this data. This data space will be used to support L2 scaling solutions with high transaction throughput.

In short, the PoS consensus mechanism provides robust security and decentralization for L2 Rollups.

Danksharding forms the data layer, providing massive data availability, thus providing a low-cost venue for Rollups to publish their data. Danksharding makes Ethereum's Rollup-centric roadmap feasible. It turns Ethereum into a settlement and data availability layer, and puts scalability in the hands of L2 Rollups. Currently, by combining Rollup and Ethereum architecture, the current transaction throughput of Ethereum of only 15-45 TPS can be expanded to up to 1000-4000 TPS. The introduction of shard chains will further expand the data storage capacity for Rollups, increasing the throughput to more than 100,000 TPS.

03. A Rollup-centric future

The shining advantage of Optimistic Rollups and ZK-Rollups is that they can solve Ethereum’s scalability problem without compromising security and decentralization. They allow for nearly instant transactions at much lower fees while reducing network congestion on the Ethereum mainnet.

However, both types of Rollups have significant drawbacks that call into question their long-term sustainability. Optimistic Rollups are subject to a week-long waiting period for exits due to their fraud proof system. And ZK-Rollups are too computationally intensive to be compatible with the EVM at this time. As mentioned earlier, developers of both solutions are working on addressing these issues.

Ethereum’s ecosystem provides ample space for the Rollup ecosystem to flourish. The Ethereum roadmap encourages experimentation, allowing the best L2 solutions to thrive in the mature Ethereum ecosystem. In addition to their respective shortcomings, future Rollup deployments must address differences in transaction costs, fragmentation of Ethereum’s sharding ecosystem, and privacy protection issues.

1) Cost

A scalable blockchain needs to be able to support an increasing number of transactions without compromising on transaction costs. As mentioned earlier, both Optimistic Rollups and ZK-Rollups solve this problem, but how do they compare?

ZK-Rollups can incur significant costs when running their validity proofs. However, as the program improves, the number of transactions included in each batch in the ZK-Rollup network will increase, which will reduce the marginal cost of each transaction. In the long run, ZK-Rollup will overcome the disadvantage of the initial fixed cost of its validity proof.

In theory, Optimistic Rollups should have almost cost-free transactions. Unlike ZK-Rollups, Optimistic Rollups do not have heavy backend computation or data compression (which further increases transaction costs). Typically, Optimistic Rollups operators bear the cost of running fraud proofs. However, publishing uncompressed data back to Ethereum increases transaction costs. In addition, because fraud proofs have not yet been fully deployed on these solutions, transaction costs on Optimistic Rollups are currently higher than expected.

On the Ethereum shard chain, the transaction costs of any Rollup solution should be significantly lower than the current situation. In theory, if ZK-Rollups can continue to reduce the marginal cost of each transaction, the cost difference between Optimistic Rollups and ZK-Rollups should be very small.

2) Fragmentation

The entire crypto ecosystem suffers from fragmentation. Most protocols operate in silos, and the industry is relatively fragmented. Moving assets between blockchains is complex and unintuitive, especially for new users.

Many Rollups networks have thriving but isolated L2 ecosystems. Developers are rapidly building applications on Arbitrum and Optimism because of the ability to easily connect components within the Ethereum protocol. However, once users put their assets into these systems, they often find it challenging to move those funds back to different platforms. "Cross-layer bridges" and "cross-chain bridges" like Hop and Connext help solve some of the interoperability problems for liquid assets. Users can seamlessly transfer assets from one platform to another relatively quickly and cost-effectively.

A cross-chain bridge locks assets in an L1 smart contract and then sends a version of the token asset to another blockchain or Rollup, or vice versa. Without EVM compatibility, users will need to give up control of their crypto assets because the security of the cross-chain bridge no longer relies on the security of the underlying blockchain. In order to transfer assets from Ethereum to Rollups that are not EVM compatible (such as Loopring or StarkNet), users can send token assets to the Rollup from a centralized trading platform, or pay Gas fees to send assets to the Rollup itself. In the latter case, these assets will exist in the L2 Rollup.

For cross-chain bridging of assets from Ethereum to Rollups, L2 bridges have declined in terms of total TVL (total value locked) since the market downturn earlier this year. Despite long withdrawal waiting periods, Optimistic Rollups like Arbitrum, Optimism, and Boba currently lead in terms of accumulated bridge TVL. zkSync is a fairly new Rollup in the space, but it is one of the few ZK-Rollups with EVM compatibility.

Similarly, in the PoS-based Ethereum world, there are issues surrounding the interoperability of Rollups. "Interoperability bridges" will become critical infrastructure components in the Ethereum ecosystem centered on Rollups. They will be committed to preventing isolated Rollup ecosystems, maintaining composability, and alleviating fragmented liquidity. These bridges are necessary means for cross-Rollup communication and asset migration.

Ethereum's Rollup-centric ecosystem encourages experimentation and evolution of current blockchain architectures. Currently, EVM compatibility limits development within the Rollup framework and restricts the freedom of smart contract innovation. EVM compatibility may become an obsolete feature, replaced by the development of future composability-driven technologies.

In the short term, Optimistic Rollups are the leading scaling solution for the Ethereum ecosystem. They have an existing advantage in EVM compatibility, which provides the required composability for their applications. However, over time, the transition to ZK-Rollups will occur naturally due to several factors: lower, trustless withdrawal waiting periods; higher throughput; superior data compression capabilities; and reduced marginal costs per transaction.

3) Privacy

Blockchain is inherently public, providing trust in the space through complete transparency. At any given time, anyone can go to Etherscan and see who is interacting with what contract and how much money is involved. While wallet addresses are an anonymous privacy solution, social media platforms provide a loophole to link public identities to corresponding wallets.

On the other hand, the current TradFi (traditional finance) system is completely private: people cannot view each other's financial status. In an era where data privacy has become a mainstream issue, it is easy to understand why retail investors, and especially businesses, are reluctant to expose their private information and link it to their financial records. The crypto space needs to be as private as the current Web2, otherwise, frankly, there will be no potential for long-term adoption.

So how exactly can developers remove identifiable information from public transactions without compromising security?

The advantage of ZK-Rollups lies in its data compression capability. As mentioned earlier, ZK-Rollups verify all transaction data off-chain and publish the verified proof and compressed transaction data back to the Ethereum mainnet. Ethereum validators do not need to interpret the data in the transaction batch, but only check and ensure that the L2 scheme has verified the data.

Privacy-focused ZK-Rollups remove identifiable information from the transaction data they bundle and publish only appropriate information. For example, Aztec created a pseudonym system in its Rollup solution to completely separate transaction data from the parties involved. Users can send or receive tokens anonymously and interact with decentralized applications through their upcoming Aztec Connect bridge.

However, this increase in privacy comes at a cost. The Rollup cannot process lists of transactions as before. Aztec introduced a second zero-knowledge proof (so now it is a ZK-ZK-Rollup) to verify a list of ZK proofs, each of which verifies a private transaction. Polygon launched their enterprise-focused privacy Rollup, Polygon Nightfall, which runs an Optimistic Rollup on top of the ZK privacy Rollup.

Previous privacy technologies, such as Monero, Zcash, and Tornado Cash, have either failed to gain mainstream adoption or made privacy opt-in for users. When investors want to make their transactions private, they must opt-in to use privacy. However, the party on the other end of the transaction may not do so, so one side of the transaction is public. Over time, dynamics in the transaction data can be pieced together to reveal the identities of both parties, defeating the purpose of hiding identities from the transaction in the first place. This vulnerability in applications that offer opt-in privacy defeats the purpose of building private transactions in the first place, and privacy should be the default setting in applications.

Enabling private transactions is simpler for ZK-Rollups because Optimistic Rollups publish all transaction data to their parent chain. However, there are options for enabling private transactions with Optimistic Rollups. They can follow Tornado Cash's lead and cut off the on-chain association between the source and destination addresses. Current Optimistic Rollups can host privacy-oriented ZK-Rollups such as Aztec as L3. Future deployments can also mix their solutions, such as Polygon Nightfall, and implement ZK-Rollup to verify private transactions.

04. Final Thoughts

Optimistic Rollups are ahead of ZK-Rollups in the current market. Since they are generally equivalent to EVM, Optimistic Rollups have the composability required for their applications and are simpler due to the lack of complex off-chain calculations. In the long run, if ZK-Rollups continue to improve their technology, they have the potential to win the L2 Rollup competition based on better product fit. As more EVM-compatible solutions emerge, ZK-Rollups should become more cost-effective, take advantage of unprecedented transaction privacy, and produce a more harmonious cross-chain and cross-layer bridging experience.

Ethereum is not the only L1 blockchain focused on a Rollup-centric roadmap. Tezos, NEAR, and Celestia have already moved to some kind of Rollup-oriented blockchain. Moreover, the current L1 competition may not disappear in the near future. These L2 ecosystems may have to compete not only with each other, but also with L1 ecosystems (such as Solana) and even L2 sidechains (such as Polygon).

Ethereum's transition to a Rollup-centric one will take several years. Current Rollups will take time to be implemented in a proper way, leaving plenty of room for implementation errors. While both Optimistic Rollups and ZK-Rollups have their own trade-offs, the possibility of a cohesive Ethereum ecosystem bodes well for further cryptocurrency adoption overall.