Opportunities and Challenges of Ethereum Layer 2 from zkSync

This article is an AMA with zkSync founder Alex Gluchowski and ChainNews Research Director Pan Zhixiong, organized by imToken.

Last Thursday, imToken invited zkSync founder Alex Gluchowski and ChainNews Research Director Pan Zhixiong to share their views on Ethereum scaling solutions, including zkSync. Alex also responded to the community's concerns regarding the zkSync Native Token mechanism and distribution plan. Below is the transcript of the conversation:

Q1: Many users have recently focused on the zkSync airdrop. How does zkSync consider the Native Token mechanism and distribution plan?

Alex: We believe that the decentralization of the initial token is crucial for any crypto protocol. zkSync intends to allocate the majority of the tokens to the public in a fair and permissionless manner. We are still exploring how to achieve this goal specifically.

Q2: What will the relationship between Layer 1 and Layer 2 look like in the future, and how will they each capture value?

Pan Zhixiong: Layer 1 will still be the foundation of security; it should carry the most important consensus, and any significant and high-value settlements should be realized through Layer 1, ensuring that security is not compromised. From this perspective, the value of ETH will still be influenced by the strength of PoW or PoS consensus. Ethereum Layer 1 will become an expensive settlement layer, secure enough while also providing sufficient scalability to prepare for Layer 2.

Layer 2 will seek a balance between security and speed through paradigm shifts. For example, Rollups will aim to increase throughput while minimizing security reductions, whereas technologies like Plasma may slightly reduce security but significantly enhance throughput. Depending on the application scenario, applications can choose the scaling solution that suits them. Therefore, for Layer 2, their methods of capturing value will also depend on their performance, security, and ecological considerations.

Q3: Between Eth2 and Layer 2 scaling solutions, which will succeed faster? What are the issues with the other solution?

Alex: When Vitalik Buterin proposed the Rollup-centric Ethereum roadmap in October 2020, he already knew that Layer 2 would arrive sooner than Eth2. This is indeed happening; it seems that most EVM-compatible Rollups will launch this summer, including zkSync 2.0.

Pan Zhixiong: Undoubtedly, Layer 2 will go live much faster than Eth2 and will become the most important scaling means. In other words, Layer 2 is not an intermediate solution but part of the ultimate solution. Eth2 has undergone significant changes after discussions among developers at the end of last year, with the core being to incorporate Rollups or Layer 2 as part of the 2.0 scaling path. Layer 2 has become an unavoidable scaling solution.

The original sharding plan has also shifted from execution sharding to data sharding, and its technical details have not yet been fully determined. What is known is that Ethereum core developers will likely advance the next steps for 2.0 this year, merging the original ETH1 blockchain with the 2.0 beacon chain, but this step will not fundamentally enhance performance. This year, Layer 2 is already capable of supporting large-scale applications, but in the long run, 2.0's data sharding can provide more possibilities for Layer 2, and the two are complementary.

Q4: With the emergence of numerous Layer 2 solutions, we will face cross-chain asset exchange issues. Do you have any preferred cross Layer 2 bridge solutions? Why?

Alex: Cross Layer 2 bridge solutions will indeed be very useful, especially for moving small assets quickly between Rollups. In my view, Connext is particularly promising, but it cannot solve the composability issues of DeFi projects. For protocols to achieve true composability in the way we are accustomed to on Layer 1, they must all reside on the same Layer 2.

Q5: From a media perspective, what challenges do domestic Layer 2 projects currently face in China? What development trends do you think the entire Layer 2 market will present?

Pan Zhixiong: Building Layer 2 networks is actually a resource-intensive endeavor, with high demands for capital, technology, and ecological construction. Therefore, Layer 2 solutions with substantial capital support and numerous application endorsements have significant inherent advantages. I believe that both in terms of quantity and quality, overseas Ethereum projects have some advantages. Once they choose one or a few specific Layer 2 solutions, domestic DeFi teams may migrate there due to the ecological agglomeration effect.

For domestic Layer 2 development teams, considering differentiated strategies is advisable. Overseas teams are mainly focused on general Layer 2 solutions, while specialized Layer 2 solutions or advantages in user experience are areas that domestic teams can explore.

We can expect at least three general Layer 2 networks to go live in the second half of this year: zkSync, Optimism, and Arbitrum. Many projects may choose to deploy on multiple Layer 2s and then adjust their focus based on user development.

Q6: Why has Optimistic Rollup not yet launched on the mainnet, while ZK Rollup (zkSync, StarkWare) has been running on the mainnet for a long time?

Alex: People often underestimate the complexity of Optimistic Rollups. ZK Rollups are secure when run by a single validator, while the security of Optimistic Rollups requires multiple reliable validators. Personally, I would only invest a significant amount of funds in a completely new architecture after it has been running for at least six months to ensure there are no fundamental design flaws.

Pan Zhixiong: To clarify, the Optimism network has actually been on the Ethereum mainnet, but it is limited access, open only to certain applications, so it does not yet qualify as a permissionless network.

The scaling solutions of zkSync and StarkWare were applied early to specific scenarios, allowing them to go live on the Ethereum mainnet sooner. The overall technical difficulty of ZK Rollup is high, but ZK Rollup solutions aimed at specific application scenarios are not very difficult. For example, zkSync 1.0 only supported transfers and collaborated with Gitcoin to reduce costs for users making donation transfers. StarkWare initially partnered with DeversiFi to launch a decentralized exchange, focusing on trading scenarios.

In the future, both zkSync and StarkWare will gradually evolve into general Layer 2 networks where any application can be deployed on their networks.

Q7: What was the reason for Matter Labs to choose the Layer 2 track and develop zkSync?

Alex: We realized that Ethereum would soon scale, and it would be meaningless to build another competing Layer 1. Those who invest heavily on-chain will always choose the platform with the highest security. In this regard, the Ethereum network is second only to Bitcoin. For the same reason, we believe that everyone will ultimately choose ZK Rollup because it is the only Layer 2 solution that offers the same security guarantees as the mainnet.

Q8: Submitting batch transactions by centralized orderers is very efficient, but it also leads to a problem—single point of failure. How do you plan to address this in the future?

Alex: There will be a single point of failure, so we are researching Layer 2 consensus mechanisms to make validators more decentralized.

Q9: What are the challenges of zkEVM? Recently, both zkSync and StarkWare announced support for general smart contracts on ZK Rollup. Why did this not happen earlier?

Alex: Porting smart contracts developed for EVM to ZK Rollup is more challenging than porting general computation. The reason is that EVM operations are heterogeneous, meaning the workload they require is variable, unlike general computation (like payments, transactions, etc.). Therefore, it is difficult to place smart contracts into a fixed zero-knowledge circuit structure without incurring significant additional costs.

The key to solving this problem for us was using PLONK recursion, which we first attempted in the summer of 2020. Recursion allows us to combine specific circuit structures optimized for different EVM operations. StarkWare cannot be compatible with EVM because heterogeneous recursion is currently impossible in STARKs.

Pan Zhixiong: Using zero-knowledge proofs in general computation should be an emerging field, and there are not many teams globally capable of producing production-level tools, with only a few like Matter Labs and StarkWare. Zero-knowledge and ordinary software development are entirely different; it has a unique system they refer to as circuit design. Because it does not just perform computations but adds privacy or anonymity features, some costs must be incurred, and specific technical details may require particularly experienced experts to discuss.

The absence of this earlier was partly due to the high difficulty and lack of financial support. The rapid development of ZK Rollup technology in recent years is also due to the heavy reliance of crypto assets, Bitcoin, blockchain, Ethereum, and other technologies on advancements in cryptography, with zero-knowledge being one of the technologies that can solve specific problems.

Q10: Currently, the cost of deposit and withdrawal operations on zkSync is relatively high. How does zkSync view this situation? What usage scenarios do users have on zkSync?

Alex: Operations on Layer 1 are inherently expensive. Our planned solution is to integrate directly with exchanges and OTC services, so there will be no need to go to Layer 1 at all. We are currently working on these integrations.

Q11: At the end of March, zkSync released the 2.0 Roadmap. What are the highlights of the zkSync 2.0 upgrade that users can look forward to?

Alex:

1. EVM-compatible smart contracts will appear on ZK Rollup!
2. Optional ultra-low-cost zkPorter accounts. In my opinion, this is the biggest breakthrough in Layer 2 since the invention of Rollup. We will soon publish an article detailing this matter.

Pan Zhixiong: This is the first production-level general programming technology supported by ZK Rollup technology, which is itself epoch-making. The difficulty of ZK Rollup technology is high, and the fact that Matter Labs can support it in such a short time demonstrates their solid engineering capabilities and cryptographic technology. For users, what they need most is low-cost access to a permissionless open network, whether for financial services, gaming, Web3 applications, storage, etc., all of which can improve efficiency without compromising security.

We look forward to wallets, including imToken, quickly following up on user experience upgrades, allowing users to use Layer 2 networks more seamlessly. Developers can also quickly deploy Layer 2 versions, enabling users to experience low-cost and more efficient blockchain technology. Finally, various service providers can offer comprehensive services, reducing the cost of entering Layer 2 and providing a more seamless experience, such as exchanges and liquidity providers supporting Layer 2 networks to create a closed-loop user experience.