What to expect? This report will provide an overview of zkSync and its key characteristics. Also included is an assessment of its potential competitive advantage, an overview of account abstraction, and an analysis of the L3 future envisioned by zkSync.
What is zkSync?
zkSync is a general-purpose smart contract Validity Rollup platform. It is focused on aligning with the Ethereum community and is widely expected to be the first EVM-compatible Validity Rollup on mainnet.
Where is zkSync Today?
- On October 28th zkSync launched their ‘Baby Alpha’ which is limited to the team using the mainnet.
- Following this, the ‘Fair Onboarding Alpha’ will be limited to projects who have registered to participate. This means that they will be able to test out their products for any bugs or issues before real users come in. All code will be open-sourced at this stage.
- Finally, the Full Alpha launch will occur (expected Q4 2022 or early Q1 2023).
This step-by-step approach makes sense as a botched launch would likely have serious negative consequences not just for zkSync but for Ethereum scaling projects (almost all of which are not close to decentralization).
Matter Labs have raised a total of $458m, with the most recent being a $200m Series C round co-led by Blockchain Capital and Dragonfly. Previous raises included a $200m ecosystem fund, a $50m Series B led by a16z and $8m in Series A and Seed.
This makes zkSync extremely well capitalized and a force to be reckoned with in the L2 wars.
- Starkware: $273m total funding
- Fuel: $81.5m total funding
Funding rounds do not mean success, but zkSync has raised very large funds to continue executing on its vision for a number of years. zkSync is expected to become the first live general-purpose Validity Rollup with Ethereum compatibility, which could give it a significant advantage. In addition, its hard stance on aligning with the Ethereum community could help it bootstrap a strong ecosystem and competitive moat that other rollups may struggle to overcome as they arrive on the scene with their mainnet deployments.
So what are some of its key characteristics?
zkSync has positioned itself as an Ethereum-focused scaling solution. It has aligned itself as much as possible with the Ethereum community. Most rollups have aligned themselves with Ethereum because it has the greatest developer share and user activity. However, some remain open to the possibility of deploying on other chains and in different configurations (a view that makes sense; why unnecessarily constrain yourself to one network?).
zkSync is fully focused on Ethereum, and prides itself on the 5 key qualities that demonstrate its alignment with the Ethereum community:
- General purpose
- Supports Solidity
- Gas in ETH
- Open Source
This demonstrates zkSync’s focus on aligning with Ethereum. The network effects of the EVM are obviously very strong, so this tactic makes sense - at least in the short term. However, an argument against this approach is that the EVM has a number of design constraints, and other virtual machines and programming languages are better suited to the applications of tomorrow that will help bring crypto mainstream. This is the approach that other scaling solutions such as Starkware and Fuel are taking. However, the wealth of infrastructure around the EVM is powerful, and it is no coincidence that 9/10 top chains by TVL are EVM-compatible.
- zkSync achieves EVM compatibility by having a VM that adapts EVM development tools and works well with validity proofs. It is not fully EVM equivalent, as Matter Labs have opted for a design that can more easily work with validity proofs.
- This design choice reduces the burden on Matter Labs to build a system that is compatible with all EVM opcodes which is far more complicated (e.g. Scroll).
- This means that some adaptation might be required for dApps porting to zkSync than would be the case with a more EVM-equivalent design.
Different Validity Rollup Designs
Nansen published an article on different Validity Rollup designs here. At a high level zkSync’s EVM compatibility occurs at the language level, whereas Hermez and Scroll are going a step further by building a VM that can interpret EVM bytecode (an approach that is notably more complex to build and was long considered infeasible).
By deploying at the language level, zkSync does not need validity proofs throughout the steps of EVM execution which should make decentralizing the proving process easier. Proponents of a full zkEVM argue that a bytecode-level zkEVM is the gold standard in terms of security.
Deploying on zkSync
- By not implementing EVM at the bytecode level, zkSync arguably adds an additional risk (due to the differences between the two environments). However, it remains to be seen if/how problematic this is. zkSync claims that 99% of code can be deployed directly to zkSync 2.0 without any need for alterations. It will be interesting to see how this transpires in practice.
- zkSync has pledged to open source everything this year, demonstrating its commitment to open source and Ethereum’s values. zkSync can be described as EVM-compatible at the language level. This means that the zkVM compiles EVM language e.g. Solidity to a SNARK-friendly VM. This adds some compiler risk to the protocol.
Token utility is a problematic topic. There is little consensus as to how to incorporate satisfactory value accrual mechanism(s) into a token to give it fundamental value. This is due to the desire for projects to avoid falling within the remit of security regulations.
For example, protocols such as Uniswap - arguably the golden child of DeFi, has a token that has uncertain fundamental value. While it is used for governance and rewards, it has no real value accrual properties that give it long-term fundamental value (other than governance (which is itself of dubious value)).
Layer-2 tokenomics have the additional caveat of being built atop another protocol with tokenomics (e.g. ETH). Adding an additional token on top runs the risk of adding additional risk and friction points. However, it is necessary to provide incentives for growing out and maintaining the ecosystem.
- zkSync has opted to have gas fees paid in ETH - as a strong signal of allegiance to Ethereum and the Ethereum community. While some rollups may just be appeasing the Ethereum community without any real loyalty, having ETH as the gas token is a strong statement of genuine intent.
- Having the gas token as ETH also helps avoid additional friction (given that transactions need to be settled in ETH at the base layer regardless). Having a native token as a fee token is likely to harm the user experience. Ultimately, the user experience will win out in an increasingly competitive rollup space.
- Beyond this, Matter Labs have remained quiet on the utility of the token. It will be interesting to see what they devise - as the option of a fee token (which has obvious value accrual properties) is off the table.
- Governance on rollups is also potentially problematic as this introduces the additional risk of governance capture. Most tokens are de facto centralized in terms of ownership - with a few actors capable of influencing governance decisions. This adds risk.
- Taking the approach of a typical PoS token would introduce an additional element of censorship risk at the Layer-2 level and is arguably undesirable as a result.
- The team has claimed that 2/3 of the tokens will be allocated to the community/ecosystem, while 1/3 will go towards incentivising the team and investors. This is a healthy split, as many tokens are overweight in favour of insiders, while ⅓ of tokens should provide ample incentives for contributors. Starkware, for example, has 50% of the token supply allocated for insiders, and zkSync CPO Steve Newcomb has suggested 50% of Arbitrum’s token supply is also for insiders.
How Strong is First-Mover Advantage?
While being first to market will undoubtedly give zkSync an initial advantage, how strong and sustainable is this? To explore this, it is worth examining EVM-compatible chains.
In the 2020/21 bull run, there was a meteoric rise of EVM-compatible chains as they provided alternatives to users from Ethereum mainnet which was prohibitively expensive.
This gave rise to players such as:
- BNB Chain
All these chains grew extremely quickly. However, the ease of deployment across EVM chains made developers and users very mercenary - quickly moving from one chain to the next without any loyalty. Taking BNB Chain as an example; while it had first mover advantage, there are few projects today that can be said to have enduring value.
Will zkSync be Vulnerable to Mercenary Developers/Users?
zkSync may be different as it is strongly aligned with the Ethereum community. It is widely regarded that Validity Rollups are one of the most promising technologies for scaling, and zkSync's alignment with Ethereum, especially with ETH as the gas token, may leave developers with a disposition towards building there. In fact, 150 Ethereum projects have stated their intent to deploy on zkSync at launch. This is the largest of any L2 at launch and includes bluechips such as Aave, Uniswap, Chainlink, Curve and others.
However, zkSync’s long-term success will depend on the designs and UX of other EVM-compatible Validity Rollups being built e.g. Scroll. If these can launch with an objectively better product, then the first-mover advantage could be pretty meaningless. However, if zkSync and its L3 ecosystem can generate strong traction before viable competitors launch, its first-mover status may give it a significant competitive advantage.
Account abstraction refers to the aim of reducing the two account types on a blockchain (wallet addresses and contracts) to one (contract accounts) - reducing complexity for users who will no longer need to make a distinction between the different account types.
Ethereum has been looking to implement it since 2017, however, it is a very challenging undertaking that has been sidelined in favour of the scaling roadmap. zkSync, (and other L2s such as Starkware and Fuel) however, will incorporate account abstraction by default. Essentially, accounts can have arbitrary logic implemented into them.
So what? While this may not seem exciting on the face of it, it actually opens up some really interesting UX improvements.
Approve Multiple Transactions at Once
Using dApps on Ethereum is frustrating, as a transaction is required (and gas fees) for every on-chain interaction. This is terrible UX. With account abstraction, multiple transactions can be bundled together. The following diagram shows how this greatly improves the user experience:
One of the greatest UX issues in crypto has been seed phrases and the total consequences of forgetting your seed phrase.
Social recovery enables users who lose their private key can authorize a new wallet as the legitimate owner.
This may be:
- Your hardware wallet
- A trusted friend/relative
- A third-party service
- Or even a combination of the above.
With social recovery, you still have self-custody of your funds and are in control of your assets. Vitalik is a strong advocate for social recovery wallets, stating that they are his ‘preferred method’ for securing a wallet.
Account abstraction enables users to set their account to require signatures from multiple keys, with transactions only executed if specified conditions are met. While this sounds similar to using a multisig like Gnosis safe, account abstraction can give wallets far greater customizability and usability than a typical multisig.
This is a good demonstration of some of the customisation that account abstraction can offer users in improving the user experience. The level of security can be tailored to the user’s needs.
Fees In Any Token
AA can enable the payment of gas fees in any token. While it is uncertain if zkSync will enable this (as they have expressly specified gas fees will be in ETH), it could hypothetically operate in the background with the fee token swapped for ETH.
This enables wallets to pre-approve certain rules for interacting with dApps so that you can use it as frequently as you like without having to sign transactions. This is especially applicable to blockchain games.
- With this, dApps can be freely used within specified boundaries, enabling users to maximise their ease of use while maintaining boundaries (and in effect minimizing risk).
- At present, users need to trust an entity to sign things on their behalf (or sign everything themselves).
AA enables plugins - meaning that users can add and remove functionalities to the account after it has been created.
Another potential feature of account abstraction is time-bound transactions - where the user can create a transaction that executes at a specified future time.
Useful resources to understand account abstraction:
zkSync also aims to operate as a volition - labeled as zkPorter. zkPorter allows users to choose between Validity Rollup (on-chain data availability) and Validium (off-chain data availability). This allows users/developers to select the security they need for their specific applications. For example, a gaming application may opt for a Validium as the additional cost of Validity Rollup security is not necessary, whereas a DeFi application will likely opt for Validity Rollup due to its enhanced security and on-chain data availability.
zkPorter off-chain data will be sent to the Guardian Network - which will be a PoS network secured by zkSync’s (yet to be released) token. This adds the trust assumption of ⅔ honest validators and demonstrates the tradeoff between Validium and Validity Rollup - while Validium is far cheaper, it has lesser security.
What are L3s?
To understand what is meant by ‘Layer-3’, it is necessary to understand L2 rollups. L2 rollups refer to blockchain scaling solutions that handle transactions off Ethereum L1, with settlement and data availability on L1. L3 is to L2 what L2 is to L1.
- L3s enable customized scaling for specific applications that want computation outside that of the EVM.
- The general consensus is that L3s enable customized functionality for applications, and leverage L2 for scale e.g. privacy.
- L3 is suited to Validiums which may be suited to application-specific use cases e.g. gaming, and enterprise applications.
L3 puts state and execution on multiple servers, and is arguably necessary for compute-intensive applications. Expecte benefits include:
- Solutions to prevent MEV potentially resulting in little to no MEV.
- Can be optimized to the application
- Fee markets
- Hardware requirements
Fractal hyperchains is the term that Matter Labs are using to describe their vision of L3s on zkSync. All Fractal Hyperchains will be bound by the same circuit technology and proven by the same prover. The significance of this is:
- By having the same prover, Hyperchains have a native bridge between each other. Bridges have been key sources of failure in the ecosystem, and a native, trustless bridge will be key to enabling scalability, safe composability and interoperability. Matter Labs have suggested that in the future there will be no bridge at all between hyperchains. However, this is purely hypothetical. Although there are claims of ‘limitless’ performance etc - it is unlikely that any single solution can scale infinitely, so it will be interesting to observe how much scalability L3s will enable when they launch.
- L3s will be highly customizable, with many app-specific chains envisioned.
- L3s will be able to choose from 3 data availability options depending on their requirements:
- Validity Rollup (most security, most expensive)
- Validium (data availability off-chain, fastest and least expensive)
- Volition (ability for the application/user to choose between Validity Rollup and Validium for their transactions).
- zkSync expects to launch its L3 proof-of-concept in Q1 2023 and claims a number of big brands are looking to deploy their own L3.
- Matter Labs also states that their LLVM compiler can support any modern programming language e.g. Solidity, Rust, C++ etc, which will open up the space. It will be interesting to see how effective LLVM is in compiling arbitrary coding languages to produce high-quality and secure dApps.
The Importance of Scalability and L3s
While the purpose of zkSync is to scale Ethereum, it is somewhat ironic that at present, there is a huge amount of unused block space across many of the existing blockchains. This is because there are no applications that have attracted mass adoption. The potential scalability and customization made possible by L3s could pave the way for the mass adoption of web3 technologies. Ultimately, the success of these scaling platforms will depend on whether or not applications suitable for mass adoption can be developed.
Main Battleground For Validity Rollup Wars Could Be The Prover
While there are a number of Validity Rollup solutions competing to launch their mainnet product and build the best ecosystem, Matter Lab’s Chief Product Officer has argued that the key focus should be on which project can build the best prover.
In order to realise the L3 hyperchain vision, zkSync’s prover will need to be accepted by the industry as the best. If this happens, it could facilitate a large inflow of builders seeking to deploy on L3 in a composable and interoperable network enabled by the shared prover.
- If projects/networks use the same prover, they will be able to interoperate with each other (secured by cryptography (theoretically)). The vision is that most on-chain activity will leverage the same prover which will enable enormous advancements in the potential for Web3 technologies.
- On the Bankless podcast, Matter Labs’ Chief Product Officer used the analogy of SSL to describe the potential of zkSync. SSL stands for Secure Sockets Layer and is a protocol for browsers and servers for authentication, encryption and decryption of data on the internet. Before the adoption of SSL, e-commerce was niche and many did not trust putting their credit card details online. However, following the widespread adoption of SSL and the confidence in the technology, e-commerce was enabled to grow and flourish.
If zkSync’s prover can be accepted as a universal standard that is safe and secure, the vision of a vast network of securely interoperable Fractal Hyperchains may be possible. This of course also depends on whether L3s and blockchain technology itself can achieve global scale product market fit.
State and Decentralization
Ethereum will move from a bandwidth-constrained environment to a bandwidth-abundant environment with the implementation of proto-danksharding and danksharding. Nansen’s report here explains the Ethereum scaling roadmap.
There is an argument that rollups today are overly focused on optimizing for data rather than bandwidth, leading to issues around growing state. This leads to challenges around decentralizing the prover/sequencer in the future.
L3 / Fractal Hyperchains may provide a satisfactory solution to reducing state growth. However, this remains to be seen and could be a problem that occurs at the L2 level in the future if they fulfil their potential of hosting very high volumes of on-chain activity.
Decentralization is at the end of the roadmap for many rollups, but it is uncertain how this will be achieved, and how quickly it can be done.
Validity Rollups can be composable with the settlement layer e.g. Ethereum L1. In order for this to occur, a validity proof would need to be generated for every block. Of course, Validity Rollup technology is not at this level; although, the rate at which it is evolving has even surprised Vitalik Buterin. This would enable L3s to compose with the underlying L2.
- However, this would require waiting for the next block to compose back. This should be a non-issue for most applications. One way of solving this is that L2 offers L3 pre-confirmations on transactions which should make L3 and L2 transactions atomically composable.
- Ultimately, it is possible that there can be n number of L3s composing to the same unified state and a single succinct validity proof verifies all L3s. This would result in full composability.
Polynya has argued that this is a viable path forward in terms of maximising for security, decentralization, and scalability. Read Polynya’s thoughts here.
This is exciting, however, it remains to be seen how these rollups / L3 will operate in practice. zkSync is leading the way, and will likely provide the first example of L3 in action.
- zkSync’s latest funding round of $200m and total ecosystem funding of $485m make it one of the best-capitalized L1/L2s with an exciting scaling roadmap.
- Its alignment with Ethereum is notable, demonstrated by gas fees paid in ETH. While this will appeal to Ethereum builders, it is unclear how powerful their first mover advantage will be as a general-purpose EVM-compatible Validity Rollup.
- Native account abstraction can help significantly improve the user experience, potentially opening up crypto to mainstream use. It will be interesting to see what account abstraction features will be added in the short-to-medium term and if they can live up to their promise.
- L3 will be a key battleground for ZK-Rollups, and zkSync is seeking to become the accepted standard for hosting L3s which can be securely interoperable with one another through cryptography. Their L3 PoC will be live in Q1 2023 which is something to pay close attention to. It will be interesting to see to what extent L3s can be built with arbitrarily different design mechanisms and how customizable L3 will be.
- A key challenge (of which there has not been a clear solution to) will be decentralizing the prover. Most rollups have decentralization on their roadmap, but doing so in a way that is anti-fragile and ensures continued high performance remains to be seen.