What makes zkSync different?

br1L...uJ3h

11 Jan 2024

142

Alex Gluchowski is the CEO, Co-Founder, and “face” of Matter Labs — the development team behind zkSync — an upstart zero-knowledge Ethereum rollup/L2.

The German-based developer group is recognized as one of the early proponents of Ethereum rollups — having launched the very first zero knowledge prototype in 2019 shortly after establishing the development team.
Per Gluchowski, the thesis was (and still is) that Ethereum would require scaling solutions in the near term. The team concluded that entities putting lots of value onchain will always pick the platform that offers the highest security — with Ethereum being #2 in this category (behind only Bitcoin). For this reason, it didn’t make sense to build a competing L1. Rather, the team decided that the next wave of app developers will seek a rollup such as zkSync — since it offers the same security guarantees as Ethereum L1 with the added benefits of privacy, lower costs, and higher throughput.

Having launched multiple versions of the protocol, conducted robust testing, several internal and external security audits, public contests, and bug bounties — the Matter Labs team launched its mainnet to the public on March 24, 2023 (just 9.5 months ago).
The team is now up to 116 employees (per LinkedIn) and has become the fastest-growing Ethereum rollup over the last 6 months.
Let’s go onchain to check the progress to date:

Core developers & transaction counts

zkSync is #1 in terms of core developers when compared to Arbitrum, Base, OP Mainnet and Starknet — it’s main competitors.
They are also #1 in terms of transaction counts over the last 180 days.

Takeaway: It’s really hard to attract and inspire a talented team of core developers to build a blockchain. It’s even harder to get other developers to come in and build apps “on top.” And it’s even harder to get users to come in and use those apps.
This data tells us that zkSync is seeing early success on all three fronts. Stay tuned as we look at user retention figures later in the report.

Bridged assets

The total value deposited to zkSync’s bridging contracts is currently $591m. With that said, in total, the L2 has only $167m of value locked within project smart contracts — with SyncSwap the leader at $46m today (per defillama).

Active users

Below we can visualize zkSyncs active users vs Arbitrum, Optimism (L2), Ethereum, Solana, and Avalanche (L1s):
As we can see, despite its nascent stage, zkSync is currently the leader amongst the heavyweights. Impressive indeed.

Product: What makes zkSync different?

As ever, our goal is to keep these sections as high-level as possible — so that you can stay out of the technical weeds, yet come away with clear takeaways.
To provide a quick primer — Ethereum rollups (L2s) are designed to scale the network with faster throughput and lower costs by batching transactions off-chain — and then anchoring proofs of the data on the L1. In doing so, rollups inherit the security guarantees of the more decentralized and secure L1 — while solving the scalability constraints of the base layer. This then enables more interesting use cases for developers to build consumer applications — leveraging the rollup/L2 for execution, and the L1 for security/settlement/data availability.
Our recent coverage of Ethereum L2s (Base, Arbitrum, and Optmism), has been largely focused on optimistic rollups.
zkSync is taking a different approach toward scaling Ethereum. They are doing it with a zero-knowledge rollup.
Below are some of the primary differences between zero-knowledge rollups and optimistic rollups.
Security guarantees: in both cases, security is derived from Ethereum because L2 rollups offer trustless bridges and do not have their own consensus — which differs from side chains such as Polygon (which has its own validator set and uses the MATIC token to pay for transactions).
Transaction verification: Optimistic rollups use “fraud-proofs.” In essence, every transaction “optimistically” assumes all transactions are valid. However, there is a 7-day “fraud-proof” period where anyone can challenge the validity of the transaction, and is economically incentivized to do so.
Zero-knowledge rollups use “validity proofs” instead.
Timing of transaction verification: 7 days for optimistic rollups vs nearly instant for zero-knowledge rollups. Why? There is no “fraud-proof” period for zero-knowledge rollups.
Privacy of data: As the name “zero-knowledge” would imply, when zk rollups such as zkSync anchor proofs of the data to Ethereum, only the “proof” is visible — and not the transaction details. With that said, transactions initiated on zkSync (at the L2 execution level) are transparent, just like Ethereum today. Anyone can see the sender, the recipient, and all the details of the transaction through a zkSync transaction explorer. Per the protocol’s governance docs, the team is focused on building out scalability first and will be implementing private transactions later.
Efficiency of transactions: In terms of the gas/data costs, zero-knowledge rollups use less gas at the L1 level since only the proof of the data is sent. Optimistic rollups post the entire transaction data to the L1 — making them more costly.
Capital efficiency: Because there is no 7-day fraud-proof period, users can withdraw funds almost immediately with zk proofs vs the significant wait for optimistic rollups today.
Transaction costs: Zk rollups cost more at the L2 level since they are more complex/technical — making the computation more expensive than optimistic rollups today.
Complexity to build: From a technical/math perspective, zk rollups are more difficult to build than optimistic rollups today.
Takeaways:

Both technologies are relevant.
Strong teams are building zk solutions as well as optimistic solutions.
We expect consensus to eventually form amongst app developers in terms of which types of applications should be built on zero-knowledge rollups vs optimistic rollups.
We haven’t seen anything to indicate clear consensus has formed in the market just yet — an indication that we are still in the sandbox/experimentation phase of innovation.
Depending on who you speak to, you’ll hear differing opinions from various developers and engineers as to which technology will be more important in the years to come.
At the end of the day, successful teams will out-execute their competition over a very long period. In some cases, optimistic rollups will incorporate zk rollups and vice versa — as we are already seeing with Optimism. More on that in the section that follows.

Let’s go back onchain to see which applications are driving the most gas fees, transactions, and users on zkSync.

Top projects by 180-day gas fees

SyncSwap is the largest DEX with $56.9m in value locked currently and about $3-5m in daily trading volume. The protocol has stiff competition within the ecosystem as Maverick Protocol, Mute.io, iZUMI Finance, SpaceFi, Pancakeswap, Odos, and Velocore are all competing to control the DEX market on zkSync.
Circle is the company behind USDC - the second-largest dollar-backed stablecoin.
Dmail Network is an AI-powered decentralized communication infrastructure built to provide encrypted emails, unified notifications, and targeted marketing across multiple chains. It has 4.82 million users to date.
Reactor Fusion and Era Lend are nascent decentralized lending protocols.

Takeaway:
zkSync is a DeFi chain right now. We think this makes sense as DeFi protocols are typically the first to market on new blockchains — creating foundational infrastructure for the build-out of the rest of the ecosystem. We expect to hear about many new projects building on zkSync this year.
One challenge for DeFi in its current onchain state is lack of privacy — which makes it impossible for institutions to enter the space today. If zkSync can make all transactions private, we think it could give them a first-mover advantage for financial use cases.
We should also note that Optimism (optimistic rollup) is currently moving forward with two proposals to add zk-proofs natively — which would revamp the tech stack and add privacy to the OP Mainnet. We understand that Coinbase is pushing for this as Base is currently built on the OP stack. Per Jesse Pollak (protocols lead at Coinbase) expects to see multiple implementations of zk rollups included in the OP stack in 2024 and has hinted that starting with optimistic implementations was simply the path of least resistance initially.
As ever, these systems continue to evolve and expand.
What we see today will be different tomorrow. The teams that can consistently iterate, pivot, and execute will ultimately be the long-term winners in our opinion.

Top projects by 180-day active users

zkSync Name Service is the only new protocol when compared to gas fees above. Similar to Ethereum Name Service — the protocol enables human-readable wallet addresses/web3 domains.
Note that ENS is building a general-purpose layer 2 bridge that makes cross-chain interoperability possible for both ENS and other applications that need to be able to retrieve data from a variety of sources (including Optimism, Abritrum, Starknet, zkSync, etc). Therefore, it’s unclear if a naming service such as zkSync Name Service will be needed long-term.

Top projects by 180-day transactions

Transaction volumes on zkSync are high. For reference, Uniswap did 22.2 million transactions on Ethereum over the last 180 days (compared to 18.9m for Syncswap on zkSync).
Circle did 4.2m transactions on Ethereum over the same period and 8.4m on Arbitrum and Optimism combined (compared to 18.8m on zkSync!).

User Retention

We’ve seen some L2s gain users due to token incentives and airdrops — which is not the same as organic product/market fit.
Does zkSync fall into this category?
Given that zkSync does not have a token just yet, many users will “farm for airdrops” by bridging assets and engaging with various projects. We can typically identify these “mercenary users” in the cohort data available via Token Terminal Pro.
Let’s have a look at user retention cohorts over the last 10 months:
The data here reveals a pretty sticky user base for those onboarded at the beginning of 2023. A quick rundown:

281k users onboarded in February with 31% remaining as monthly active users in December.
552k users onboarded in March with 30.7% still using zkSync through December.
1 million users were onboarded in April with the network retaining 20.7% through year end.

Between 100 and 700k users have been onboarded each month since April. Retention rates are now dropping slightly compared to earlier in 2023, but these are still impressive figures — especially when we consider the competition.
Speaking of competition. Here’s Arbitrum (optimistic rollup) — which retained only 2.5% of its users over the same 10-month period.
Next we have OP Mainnet (optimistic rollup) which had 3.4% of its users still interacting with its smart contracts 10 months after onboarding.
Up next is Starknet (a zero-knowledge Ethereum rollup) — with 16.9% user retention in December from the February 2023 cohort. With that said, user retention is dropping with each subsequent onboarding month.
Finally, below we have Ethereum — with just 1.8% of users retained as they move to L2. Keep in mind that even when users leave Ethereum for the L2, their economic activity is still ultimately anchored to the L1. At least for now as new data availability solutions could change this in the future (a topic we’ll explore further in the Q4 update of The Ethereum Investment Framework).
Summary: 10-month user retention