ZK-Rollups Transform the Trilemma to Dilemma

Compatibility vs. Performance Ignites the zkWars on Ethereum

Article by tommasogualtieri.eth Edited by Trewkat & Hiro Kennelly Cover Art by Chameleon

This is undoubtedly one of the key moments in the recent history of blockchain, and in particular for the Ethereum protocol. Ethereum developers, like experienced gamers, are one step away from defeating the final boss: the mythical and fearsome Trilemma, and reaching the final goal: mass adoption.

Blockchain Trilemma. Source: Vitalik.ca

In general, it has not been possible to achieve network security and decentralization, while at the same time enabling scalability. This challenge has been labeled the trilemma, and it’s proven to be one of the most difficult problems to solve in blockchain architecture.

A network’s decentralization and security are related to the number of nodes over which the ledger is distributed  —  the more nodes, the more decentralized and secure the blockchain is  —  but decentralization and security are achieved by sacrificing scalability, which is the capacity of a blockchain to increase the number of transactions per second. Scalability suffers because during busy periods the process of validating transactions by every node slows down the network.

Sometimes, it is preferred to sacrifice security for scalability, such as with Bitcoin’s Layer 2 Lightning Network.

Even before the transition to proof-of-stake, the Ethereum network could process an average of 15–45 transactions per second (TPS), which may be faster than Bitcoin, but it’s not enough throughput to create a good experience for most Ethereum users, who sometimes face network congestion, high fees, and long wait times.

The solution to these problems is network scalability.

How Blockchain Networks Scale

Generally, one can scale a blockchain either at the network level or through Layer 2 scaling solutions.

For network-level scaling, these solutions can range from increasing the block size, like Bitcoin, to Ethereum’s much-discussed sharding, which would allow for the parallel and simultaneous processing of a very large number of transactions. These solutions may be effective, but they are not easy to implement.

The second, and easier way, to scale a blockchain is to employ a variety of scaling solutions. These can range from the creation of side chains, like Polygon, to Plasma chains (which have largely fallen out of favor), to State Channels, which is almost like an escrow-based solution and the one employed by the Lightning Network.

But the most promising scaling solutions may be Rollups, and in particular, Optimistic Rollups and ZK-Rollups, solutions in which Layer 2 transactions are essentially packaged and sent to Ethereum Mainnet in a bundle.

Optimistic and ZK (zero-knowledge) Rollups share the dual characteristic of processing transactions off-chain and storing the output of these transactions on Mainnet, but there are substantial differences between these scaling solutions.

Optimistic Rollups assume that all transactions are valid and legitimate ‘a priori’, and batches of transactions are sent to Mainnet without performing any compression or validity tests. In contrast, the ZK-Rollups ensure the transactions are legitimate before batching them to Mainnet.

Another fundamental difference, at least until now, was that the ZKs were not compatible with the Ethereum Virtual Machine (EVM), a circumstance that generally resulted in a preference for Optimistic Rollups. Optimistic Rollups are either EVM-equivalent, like Optimism, or EVM-compatible, like Arbitrum, and this interoperability has made Rollups the go-to scaling solution. That is, until now, as the work done by the ZK developers in recent months has been truly incredible.

https://banklesspublishing.com/media/cfc55895ef1e90172bdcdae6cbfb31ddSource: Twitter

The Rise of ZK-EVMs

It was on August 4, 2022 when Vitalik Buterin, in his article on the different types of ZK-EVMs, identified them as the ideal solution for scalability, while anticipating the need for a long and complex development process. On that occasion, Vitalik gave rise to what I refer to as ‘the Dilemma’  —  the fact that ZKs require developers to balance two competing performance metrics, as can be seen from Buterin’s diagram below:

The different types of ZK-EVMs. Source: Vitalik.ca

Compatibility with the Ethereum infrastructure and performance is measured by the ability to quickly create validity proofs, with a consequent increase of validated transactions. The two factors appear inversely correlated: the greater the compatibility with the Ethereum infrastructure the lower the performance (the ZKs that fall under level 1 in the image above) and, inversely, the lower the compatibility with the existing infrastructure, the higher the performance as measured by the speed of generating proofs, while reducing costs and centralisation risks (the ZKs that fall under level 4).

In essence, the compatibility of EVMs is a spectrum with trade-offs between high compatibility (ease of reuse from Layer 1) and high performance (speed of generating ZK evidence). The higher the compatibility with Layer 1, the lower the performance, and vice versa.

Current ZK-EVM Protocols

There are two main competitors in the ZK race, zkSync, by Matter Labs, released on Mainnet on March 24, 2023 and zkEVM, by Polygon, released on Mainnet on March 27, 2023. Their approaches represent the extremes of the above graph, levels one and four.

zkEVM is Polygon’s open-source ZK-Rollup that provides full compatibility with EVM opcodes. It is the first zero-knowledge scaling solution fully equivalent to the EVM, leveraging the power of ZK-testing to reduce transaction costs and massively increase throughput, inheriting the security of Ethereum.

Polygon’s zkEVM has been described as level 3 by Buterin but is headed in the right direction.

Polygon’s main competitor for zkEVM is zkSync. It was built by Matter Labs to be less compatible with Ethereum at launch, but to quickly execute low-fee transactions.

Alex Glukhowski, CEO of Matter Labs, speaking about zkSync, stated:

“We’re not focusing on rigidly succeeding on backwards compatibility. That is not our goal. We want to leverage the maximum potential of zk-proofs for the future,” and “We’re willing to deviate from the current standard in order to innovate faster.” 

zkSync is found at level 4 of Vitalik’s scheme (hence priority to performance). So we have two visions that might seem to be completely different, apparently at odds with each other, but which, in the end, represent two different aspects of the same goal: scalability.

In order to clear the field of doubts about performance, the Polygon team hastens to clarify that cryptographic innovations are in the works that will allow it to offer better performance than optimistics and other ZKs.


Similarly, in order to prove that lack of compatibility is not an issue, zkSync routinely states that dApps of all kinds have decided to take advantage of Matter Labs’ technology attempting to demonstrate that, perhaps, the Dilemma is not such a big deal:


Admittedly, although prioritizing EVM compatibility has significant advantages, highly EVM-compatible ZKs are often slower and require more resources than ZKs that choose a lower degree of EVM compatibility to optimize efficiency and scalability. This trade-off between performance and EVM equivalence is why zkSync chose the less-compatible approach to building its virtual machines.


So Who Is Right?

As always, the solution comes from Vitalik who concludes The different types of ZK-EVMs by saying that: 

The types are not unambiguously ‘better’ or ‘worse’ than other types. Rather, they are different points on the tradeoff space: lower-numbered types are more compatible with existing infrastructure but slower, and higher-numbered types are less compatible with existing infrastructure but faster. In general, it’s healthy for the space that all of these types are being explored.

The #zkWars have begun.

Author Bio

tommasogualtieri.eth is an attorney qualified in Italy — In house Lawyer (Commercial, Contract Law, M&A) — Blockchain, Web3 and Crypto Enthusiast — BanklessDAO Contributor — 333.builders Contributor — Member of the Blockchain Association Italy — Member of Blockchain Education Network Italia.

Editor Bios

Trewkat is a writer and editor at BanklessDAO. She’s interested in learning as much as possible about crypto and NFTs, with a particular focus on how best to communicate this knowledge to others.

Hiro Kennelly is a writer, editor, and coordinator at BanklessDAO, an Associate at Bankless Consulting, and is helping to build a grants-focused organization at DAOpunks.

Designer Bio

Chameleon is a designer and creator in the web3 space.

BanklessDAO is an education and media engine dedicated to helping individuals achieve financial independence.

This post does not contain financial advice, only educational information. By reading this article, you agree and affirm the above, as well as that you are not being solicited to make a financial decision, and that you in no way are receiving any fiduciary projection, promise, or tacit inference of your ability to achieve financial gains.

Bankless Publishing is always accepting submissions for publication. We’d love to read your work, so please submit your article here!

More Like This

Ethereum Scaling Solutions by Sandeep

The Environmental Impact of The Merge by Marathonmind

Cardano vs. Ethereum by Vi-Fi

Collect this post to permanently own it.
IndyPen CryptoMedia logo
Subscribe to IndyPen CryptoMedia and never miss a post.