Amber Group
Published in

Amber Group

Navigating Zero Knowledge


One of the most pressing challenges that the Ethereum network faces is scalability. This issue was first evident during the viral CryptoKitties in 2017 — the sensation led to a six-fold increase in total network requests, resulting in transactions taking hours to confirm. More recently, the Cambrian explosion of applications built on Ethereum, including decentralized finance (DeFi) protocols and NFT projects, has reintroduced the urgency of finding a scaling solution. Ethereum’s network only accommodates roughly 15 transactions per second (TPS), a far cry from the average 1,700 TPS that Visa processes or the average 6,000 tweets per second on Twitter. If Ethereum aims to be the dominant decentralized computing platform for the world, it needs to find a way to handle more activity in its ecosystem.

Reaching the Limits of Ethereum Transaction Throughput

Source: Etherscan

Alternative layer one protocols, such as Solana, Avalanche, and Binance Smart Chain, have gained users and developers seeking lower transaction fees, faster finality time, and/or higher transaction throughput. However, to accommodate those feats, alternative layer ones sacrifice a degree of decentralization (see: the blockchain trilemma). Furthermore, if their growth trajectory persists, these blockchains will eventually reach their throughput limits as well. For instance, the Avalanche network recently implemented an upgrade (Apricot 5) to lower fees once its network started seeing higher transaction activity — a short-term fix to what will be a long-term problem.

Enter Layer Two

A wide range of scaling solutions have been explored and launched, including state channels, plasmas, and rollups. There are several different types of layer two scaling solutions that we previously covered in this article. But rollups are viewed as the ultimate scaling solution because they enable fast and cheap transactions without sacrificing any of the security and decentralization properties of the underlying layer one blockchain. Transactions on rollups are more affordable for two reasons: 1) only a fraction of each transaction needs to be stored on the layer one blockchain, and 2) none of the computation for a transaction needs to be performed on layer one. More broadly, rollups play a central role in shifting from “monolithic” to “modular” blockchain architectures.

Shift from Monolithic to Modular Blockchains Allow for Specialization and Efficiencies

Ethereum pivoted to a “rollup-centric roadmap” last year. This month, Vitalik Buterin brought forward a proposal, termed EIP-4488, to reduce overhead costs for rollups, thus lowering L2 fees. The Merge (Ethereum’s transition to Proof-of-Stake) and sharding is expected to further reduce transaction costs for rollups.

But rollups are not solely bound to Ethereum. Arthur Breitman, founder of the Tezos blockchain, argues that rollups are the ultimate scaling solution. NEAR, another layer one, has also designed its blockchain to accommodate a modular architecture with data availability sharding. And just this month, Celestia and Evmos announced a collaboration to build Cevmos (Celestia x EVMos x CosmOS), an open, modular stack for EVM-based applications using Celestia as the data availability layer and Evmos as the settlement layer.

There are two primary types of rollups: fraud proofs (optimistic rollups) and validity proofs (zero-knowledge rollups). And while we believe that a wide range of scaling solutions will co-exist in the future, we are particularly excited about the emergence of zero-knowledge (ZK) rollups with EVM compatibility.

In this article, we dive deeper into ZK technology, explain some of the terminologies commonly used when discussing the ZK field, why the crypto community is excited about ZK rollups, and describe the current state of the market. We also provide profiles of some of the major players in the industry, providing links and references for readers interested in diving deeper.

Moon Math — What is Zero-Knowledge Proof?

The concept of zero-knowledge proofs was first introduced by MIT researchers in 1989. Zero-knowledge proofs allow one party to prove to another that a given statement is true without revealing any additional information aside from the fact that the statement is true. But the mathematics behind these proofs were considered so complex that they were colloquially termed “moon math.”

As a simple analogy, consider a circular cave with a magic door in the middle that requires a secret phrase to open it.

Alice attempts to convince Bob that she knows the secret phrase to open the door but does not actually want to reveal the phrase itself. To do so, Bob stands outside the cave while Alice goes inside. Alice takes either path A or B. Bob then enters the cave and shouts the name of the path that he wants Alice to return from.

There are two possible outcomes: either Alice knows the secret phrase and returns via the correct path, or she does not know the secret phrase and has a 50% chance of being caught lying. Of course, 50% is not good enough, so they repeat the exercise several times until Bob is sufficiently confident that Alice does indeed know the secret phrase. After just 20 times, the odds become less than one in a million.

[Note: in actual applications of ZK proofs, rules are set and defined, and as long as the rules are met, multiple iterations are not necessary for a ZK proof. For instance, in the analogy above, as long as Bob stands at the entrance of the cave, and Alice enters from one side and comes out from the other, Bob is assured that Alice knows the secret phrase.]

The above analogy demonstrates three essential properties of ZK proofs:

  • Completeness: if the statement is true (Alice knows the secret phrase), the verifier (Bob) will be convinced of this fact and accept the proof.
  • Soundness: if the statement is false, no cheating prover can convince the verifier that it is true (i.e., if Alice does not know the secret phrase, she cannot lie to Bob that she does).
  • Zero-knowledge: if the statement is true, the verifier does not learn anything aside from the fact that the statement is true (Bob does not know what the secret phrase actually is).

Applying Zero-Knowledge to Rollups

While most of the above article focuses on ZK proofs, they are somewhat of a misnomer. In fact, most ZK rollups don’t use ZKs. The magical aspect of proofs that all ZK rollups use has nothing to do with hiding additional information.

Instead, the critical property of the ZK protocols that rollups use is that it is far less costly to verify proofs than it is to compute them. Proof verification is exponentially more efficient than running the computation. This property — termed succinctness (or scalability) — provides scaling properties that the crypto community has been looking for.

For that reason, it would be more appropriate to call them “validity proofs” or “computational integrity proofs” instead. Nevertheless, due to community conventions, we continue to use the term ZK rollups/proofs for the article.

ZK rollups work by separating transaction execution from consensus and data availability. A ZK protocol cryptographically proves every batch of executions on the rollup. The proof is sent to the layer one mainnet. And because every single transaction on a rollup still stores its input data (known as “calldata”) on layer one, other nodes can verify the integrity of these transactions and decide to accept or reject them.

Hence, ZK rollups inherit the security properties of the underlying layer one protocol as nodes/miners on the L1 blockchain will reject improper transactions. For instance, double-spend is not possible on ZK rollups, since it’s impossible to create a proof of a double spend, and every state update must be accompanied by a valid proof.


The two most compelling ZK protocols in the market today are zk-SNARKs and zk-STARKs.

zk-SNARKs (Succinct Non-interactive Argument of Knowledge) were first suggested in 2012 and implemented shortly after, providing the technology a meaningful first-mover advantage in adoption. Succinctness means that time to verify scales poly-logarithmically, while non-interactiveness suggests that the proof system cannot allow any further interaction after a pre-processing phase, which may include “toxic waste” and require linear time to create.

On the other hand, zk-STARKs (Scalable Transparent Argument of Knowledge) are relatively newer technologies. Introduced by the StarkWare team in 2018, zk-STARKs have two primary advantages:

  • Transparent — the system works without a trusted set-up (i.e., it removes the “toxic waste” allowed by zk-SNARKs); and
  • Scalable — they are more scalable in terms of computational speed and size as proving time scales quasi-linearly; and, crucially, total verification + preprocessing time scales poly-logarithmically.

However, STARKs have much larger proof sizes than SNARKs. Additionally, Ethereum has a precompile for specific SNARKs, and for those SNARKs the gas cost of verifying a proof is lower than for a STARK proof.

Summary of STARKs vs. SNARKs

Source: Eli Ben-Sasson, Matter Lab’s Github repository, Elena Nadilinski

Rollups vs. Validiums

Within ZK scaling solutions, there are also significant differences between rollups and validiums. Rollups use on-chain data availability, meaning that they post data on the layer one blockchain itself. Validiums use off-chain data availability — they post validity proofs on-chain, but the data remain on a separate network.

Overview of Different Types of L2 Scaling Solutions

Validiums expand scalability. Transactional bandwidth of validiums significantly improves at ~10,000 TPS (vs. ~2,000 TPS for ZK rollups) because payment for on-chain data is not required. There are also privacy benefits from keeping data off-chain — users’ balance information is stored privately with the validium operator(s) instead of publicly on the blockchain.

However, validiums require new trust assumptions. Data availability is eventually needed to retrieve funds from the validium back to the layer one blockchain, so users must trust the operator to make the data available when needed.

Nevertheless, it is still a more secure approach than side-chains since state transitions must be verified through STARKs/SNARKs. Even assuming that the validium operator is compromised, user funds are still (theoretically) secure. Fund balances and transactions cannot be manipulated, merely frozen.

Why the Hype Now?

One of the biggest detractors against ZK rollups was the generalizability of these platforms. Many believed that creating a general-purpose EVM execution would be at least a few years out. However, breakthroughs in engineering techniques and heuristics have brought the timeline for fully EVM compatible platforms forward to 2022.

ZK rollups offer superior benefits compared to other scaling solutions across a wide range of metrics, including transaction costs, finality time, and trust assumptions. Furthermore, whereas layer one blockchains like Ethereum can be argued to have anti-network effects (experience degrades as more people use the platform), ZK rollups offer strong network effects as transaction costs are inversely proportional to network usage. Hence, many in the crypto community see ZK rollups as the ultimate rollup solution — Vitalik Buterin even suggested that “in the medium to long term ZK rollups will win out in all use cases.”

State of the Market

While ZK rollups are still a nascent field — zkEVMs have yet to fully hit mainnet — there is already an emerging ecosystem of applications, wallets, and infrastructure (note that the ZK rollup ecosystem is mainly concentrated on Ethereum, the L1 blockchain facing the greatest congestion challenges).

Nonetheless, metrics already indicate interest from users and developers. StarkWare’s StarkEx already has TVL of over $1.2 billion with 75+ million transactions executed across all deployments since mainnet launch.

StarkWare Already Has Strong Market Traction

Source: StarkWare

StarkEx’s largest customer, dYdX, is perhaps the greatest signal of product-market fit for ZK scaling solutions so far. By leveraging StarkWare’s StarkEx rollup, dYdX offers decentralized derivatives trading experience at fees that are competitive with centralized exchanges. It executed nearly $9.5 billion of trading volume at its peak, briefly surpassing Coinbase’s spot exchange trading volume.

dYdX Surpassed Kraken Futures in Trading Volume in August

Source: CoinGecko

And there are already ~212k unique addresses on zkSync 1.0. While this reflects only ~0.12% of Ethereum unique addresses, it is still an impressive feat for an early-stage platform with limited functionality and applications.

Growing Interest in zkSync

Source: Dune (@kylin)

Profiles of Major Players

StarkWare — STARK Proof Pioneers

StarkWare was founded in 2018 by scientists at the forefront of zero-knowledge computation research. As the name suggests, StarkWare’s ZK technology is built upon zk-STARKs — co-founder Eli Ben-Sasson and Michael Riabzev are the co-inventors of the STARK protocol. StarkWare’s products use a Turing complete platform and programming language for producing STARK proofs called Cairo.

StarkEx — Validium Solution with Existing Customers

StarkWare’s validium solution, StarkEx, is one of the first zero-knowledge scaling solutions that found strong product-market fit with a wide range of protocols. Both DeFi and NFT dApps have leveraged StarkEx’s scaling solution to offer a better user experience. As of this writing, StarkEx has $1.2 billion in total value locked (TVL) and has processed 65 million transactions since June 2020.

Applications Built on StarkEx

StarkNet — General Purpose Rollup

The StarkWare team is also developing a general-purpose ZK rollup called StarkNet. Its Alpha version launched on mainnet this November. It features EVM compatibility via a Solidity-to-Cairo compiler. Only whitelisted applications will be launched during the Alpha phase as the team expects to change, fix, and improve the platform. The sequencer (what executes the transactions) is also currently centralized and closed-source. However, the team plans to gradually remove the whitelist guardrails, grow the ecosystem, and open-source and decentralize the platform over time. The team terms this the “planets — constellations — universe” roadmap.

Starkware’s Roadmap to a Robust, Decentralized Ecosystem

Source: Starkware

StarkNet’s ecosystem is still relatively bare given the recent Alpha launch and the requirement to learn a new programming language. However, we expect more projects to come on board. Starkware’s team has worked closely with Nethermind to release Warp, a transpiler that converts and deploys Solidity-written smart contracts onto StarkNet. The Nethermind team plans to Warp Uniswap V3 onto StarkNet, which should pave the way for other existing projects to port their code to StarkNet and catalyze a step-change in growth for the platform.

Select StarkNet Projects

Volition — Data Availability is a Spectrum

StarkWare also pioneered a “Volition” architecture, in which users can decide whether to use validium or rollup on the L2 on a per transaction basis. Some scenarios for the Volition architecture could include:

  • A cryptocurrency trading firm starts the trading day off by moving funds to its traders’ Off-chain Data (OFFD) account. They can then trade cheaply and frequently throughout the day. At the end of the day, they can then move all funds back to an On-chain Data (OND) account to leverage layer one’s security.
  • A gamer signs up for an online game that leverages StarkEx’s (Validium) scaling solution. After a few months of playing, she earns a rare in-game NFT item with high value. Because she wants to keep the NFT item secure, she moves that item to a rollup.

Volition provides the best of both worlds — users can choose the level of security guarantees that they would like while enjoying low transaction fees.

Enabling Scalable, Secure, and Trustless Applications

Starkware has raised a total of $162 million from equity investors. This November, the Starkware team raised $50 million at a $2 billion valuation in an “opportunistic” fundraise — meaning that the company is profitable, but the team secured new funds to accelerate ecosystem and team growth.

Starkware Fundraising Profile

Figures in USD millions. Source: Pitchbook, press releases

StarkWare’s biggest strength is that it has already proven the viability of its scaling solutions. Given its proven solution, competent team, and roster of blue-chip investors across the crypto ecosystem, we expect StarkWare’s suite of rollup solutions to see strong adoption going forward.

Matter Labs — Liberty, Blockchain, and Math

Matter Labs started in 2019 and architected the zkSync scaling solution, which uses zk-SNARK technology for its rollup. The founders, Alex Gluchowski and Alex Vlasov, are both deeply technical in Ethereum and ZK research and development.

One primary difference between StarkWare and Matter Labs is the choice of scaling architecture — StarkWare uses the zk-STARK technology while Matter Labs use the PLONK zk-SNARK invented and developed by the Aztec team (see SNARK vs. STARK above).

zkSync — Open-Source Rollup

Matter Lab’s marquee product, zkSync, released its minimum viable product (zkSync 1.0) in summer 2020. A ZK rollup on Ethereum, zkSync 1.0 enabled faster payments and processed over 4 million transactions by November 2021 but offered no smart contract functionality.

zkSync 2.0 aims to fix this by introducing EVM compatibility, expanding the design space to Turing-complete operations. It will support all the developer tools used to write Ethereum smart contracts, including HardHat, ether.js, Solidity, Vyper, etc. The closed testnet for zkSync 2.0 launched in October with Curve Finance as the initial application. UniSync, a fork of Uniswap v2, is also operating on testnet to illustrate zkEVM functionality. The team claims that with zkSync 2.0, most DeFi and NFT applications will work with little to no code changes.

zkPorter — Matter Lab’s Validium Solution

Along with the launch of zkSync 2.0, the team will introduce zkPorter, a proof-of-stake chain that will allow users to post transaction data on zkPorter instead of Ethereum. Thus, zkSync 2.0 will also have a Volition design, in which users can decide between posting transaction data on Ethereum (rollup) or on zkPorter (validium).

zkPorter is still in design, but based on public information, the primary difference between zkPorter and StarkEx is that the former is account-based — a zkPorter account can only produce transactions with off-chain data availability. Furthermore, zkPorter is claimed by Matter Labs’ team to be more decentralized as data availability is secured by a network of “Guardians” incentivized by tokens. Instead, Starkware’s currently StarkEx systems relys on a “Data Availability Committee”, but StarkWare also announced a roadmap to a fully and decentralized Validium solution for Starknet.

The zkSync 2.0 Volition Architecture

Source: Matter Labs

Building an Ecosystem

Matter Labs most recently raised $50 million in a Series B fundraising round led by a16z. In addition, they opened a second financing to introduce strategic partners from all corners of the crypto industry, including exchanges, DEXs, bridges, wallets, and fiat on-ramps.

Matter Labs Fundraising Profile

Figures in USD millions. Source: Pitchbook, press releases

Investors in Matter Labs Span Across the Crypto Ecosystem

Source: Matter Labs press release

Polygon — Suite of Scaling Solutions

Polygon is perhaps most well-known for Matic, a plasma chain that evolved to a proof-of-stake side-chain. However, Polygon could be thought of as a suite of Ethereum scaling infrastructure — the company terms itself as “Ethereum’s Internet of Blockchains.” Polygon’s solutions allow developers to pick and choose the solutions they need, from execution logic to validator management.

This August, Polygon made a $1 billion bet on ZK technologies as the “best bet at solving big challenges we are facing and onboarding the first billion of users.” Shortly after, the company made a spate of announcements, including the acquisition of Hermez and Mir and the launch of Polygon Miden and Nightfall. We briefly touch on each solution by order of announcement.

Polygon Hermez — Decentralized ZK Rollup

This August, Polygon acquired Hermez in a $250 million deal, a significant premium to the market capitalization of $18 million that Hermez tokens were trading at.

Hermez is a zk-SNARK scaling solution developed by Iden3’s labs that was launched on mainnet in March 2021. Hermez created the first decentralized ZK rollup solution, in which off-chain coordinators gather requests, validate transactions, and generate SNARK proofs that get submitted to on-chain smart contract verifiers. Its value proposition lies in decentralization and efficiency. Hermez tokens (now $MATIC after the acquisition) are used in a permissionless auction system to allow network coordinators to earn the right to process transactions during a slot of time. The team believes that coordinators would be incentivized to cheaply and efficiently execute transactions by enabling a free market for rights.

Currently, Hermez is live on mainnet but only supports token transfers and is not EVM-compatible. On Polygon Hermez’s roadmap is its 2.0 version, a fully compatible Ethereum VM that will support transparent deployment of existing smart contracts without the need to adapt or transpile them. This 2.0 will also provide an L2 consensus protocol for the selection of the permissionless network operators to keep all the properties of decentralization. A proof of concept has already been built and the public testnet is expected to launch by Q2 2022.

Polygon Nightfall — Optimistic Rollup with ZK Proofs

Developed in partnership with EY, Nightfall is an enterprise-focused scaling solution that focuses on privacy. Importantly, Nightfall is not a ZK rollup but rather an optimistic rollup that uses ZK technology to enable private transactions. In its current iteration, Nightfall can perform a transfer for less than half of the cost of a public ERCx transfer while hiding the recipient and token being transferred and maintaining the security and consensus assumptions from Ethereum.

The benefits to enterprises using this solution include:

  • Maintaining privacy while complying with AML/KYC standards. For instance, when a user performs a ZK proof transaction on Nightfall, they prove that they are on a KYC-approved whitelist.
  • Providing enterprises with privacy. Enterprises seeking to use permissionless blockchain solutions still require privacy for certain data and transactions, such as sensitive customer information or proprietary datasets.

Polygon Miden — Polygon’s STARK Solution

Miden is a EVM-compatible zk-STARK rollup written in Rust. The project is led by Bobbin Threadbare, a former core ZK researcher at Facebook who led the development of Winterfell. Miden aims to guarantee “Ethereum-compatibility” by directly compiling Solidity-written smart contracts into Miden VM’s native language. It also aims to embed privacy-enabled capabilities so that it is possible to selectively reveal parts of an application while keeping the rest of the program private. And in contrast to Starkware’s StarkNet, Miden is fully open-sourced.

However, Miden is still in early development. An early prototype has been launched, but functionality remains limited. For example, loops can only be nested eight levels deep, and conditionals can only be nested at most 16 levels deep. Nevertheless, Miden’s next Github branch is active with significant upgrades to the VM to expand Miden’s feature set on the way.

Polygon Zero — Recursive ZK Proofs

Only four months after announcing the Hermez acquisition, Polygon announced another acquisition of a ZK solution, Mir, for $400 million. Mir was originally a layer one, proof-of-stake blockchain that uses ZK proofs to improve scalability by simplifying transaction execution. Mir is now Polygon Zero.

The Polygon Zero thesis is that truly scalable ZK L2s require significantly faster and more efficient proving systems. Of the ZK L2’s in production today, none have demonstrated 1,000+ TPS throughput in practice, mainly due to the computational resources required to generate proofs.

Polygon Zero has developed a recursive proof implementation that is designed to enable high-throughput ZK L2s. This proving system, called plonky2, is stated to be incredibly fast: recursive proofs take 170ms to generate on a laptop, more than 100x faster than any existing alternative.

This speed increase is important because recursive proofs allow rollups to use parallelism to improve decentralization and reduce latency. Instead of generating a single proof verifying 1,000 transactions in a batch, which would be time and compute-intensive, rollups can generate 1,000 proofs in parallel, each verifying a single transaction. These proofs can then be recursively aggregated into a single proof.

Plonky2 also allows Polygon Zero to exploit an important tradeoff in FRI, a core component of STARKs. With FRI, proofs can either be fast to generate, but big, so they’re expensive to verify on Ethereum, or slow to generate and cheap to verify. Plonky2 allows Polygon Zero to recursively verify STARKs efficiently, so transaction proofs leverage STARKs with a configuration to maximize prover speed. These transaction proofs are then aggregated into a single small proof. To bridge to Ethereum, this proof is wrapped again in a size-efficient proof, which takes around twenty seconds to generate but is much smaller and cheaper to verify on Ethereum. This allows Polygon Zero to achieve incredibly fast proofs with minimal verification cost.

Polygon Zero aims to build the most performant EVM-compatible ZK rollup. The architecture will be conceptually similar to zkSync 2.0. However, roadmap details remain scant.

Basket of Layer Twos

Co-founder Sandeep Nailwal stated that Polygon’s ambition is to become a “Layer 2 aggregator.” Indeed, the company has amassed a portfolio of scaling solutions with a wide range of architectures. And Polygon appears to have a culture of shipping and iterating on products, letting the market decide which solution it likes best. In addition, the company offers an SDK, a general-purpose data availability layer (Polygon Avail), and continues to see strong activity on its PoS sidechain. Hence Polygon could be seen as a basket of options for the future of Ethereum scalability.

Loopring — zkRollup for Trading and Payment

Loopring was launched at the end of 2019 as the first ZK rollup on Ethereum mainnet. Its first product only supported order book DEX transactions, but later versions include transfer payments and AMM swaps. This September, Loopring launched support for NFT minting, transfers, and trading. The project aims to launch an NFT marketplace by the end of Q4 2021 — given the current timing, this is likely pushed out to Q1 2022.

Similar to zkSync, Loopring uses zk-SNARK proofs for validity. But one of the main differences between Loopring and zkSync is the set of algorithms used for ZK proofs — Loopring uses Groth16 whereas zkSync uses Plonk. Although both require an initial trusted set-up, zkSync can modify and upgrade any circuit afterward while Loopring continues to need a trusted set-up.

Like all other rollups, Loopring is also working on EVM compatibility. However, Matter Labs and Starkware both achieve zkEVM by defining a language themselves (zkSync: Zinc, Starkware: Cairo) and then translating EVM bytecode into each language’s EVM code. Instead, Loopring aims to directly compile the Solidity code into the byte code of its zkEVM without any translation and directly execute the byte code in the virtual environment of zkEVM.

This November, the market speculated that Loopring was possibly collaborating with Gamestop on creating an NFT marketplace, sending the price of $LRC tokens up over 400% since initial rumors. Loopring also recently launched Counterfactual Wallet, a wallet with direct fiat on-ramp to its L2 via Ramp. It plans to expand its wallet support to other L2s, such as zkSync 2.0 and Arbitrum, as well as EVM-compatible L1s such as BSC.

Heightened Wallet and Volume Activity After Gamestop Rumors

Source: Dune (@Brecht)
Source: Dune (@Brecht)

Other ZK Platforms

Multiple other teams and projects are attempting to solve blockchain scalability using different ZK technologies and designs. In addition to other ZK rollup projects, protocols like Mina and Aleo aim to build a monolithic blockchain using ZK protocols to scale.

Other Blockchains Leveraging ZK Technology

In addition, teams like Starkware are already exploring ZK layer three scaling solutions on top of a zkEVM rollup to provide even more scalability.

Connecting Them All Together

With the range of zkEVM rollups expected to launch in the next few years, in addition to other types scaling solutions, liquidity fragmentation could become a growing issue. Fortunately, teams are working to solve this. One promising architecture is Starkware and Loopring’s proposed dAMM framework. dAMM (distributed automated market maker) is a L2 AMM that aggregates liquidity into a single L1 pool. Separate L2s can access the L1 liquidity pool asynchronously — meaning that they can access the same pool without having to communicate with each other. New smart contract logic has to be introduced and certain guardrails have to be put in place to limit impermanent loss (IL), but the novel design is a promising solution to deepening liquidity and improving capital efficiency across multiple rollups.

Proposed dAMM Framework to Enable Cross-L2 AMM

Source: StarkWare, Loopring

Bridges like Hop Protocol, Connext, and cBridge are also working to ease cross-L2 transfers. Tokemak, a decentralized market making protocol, announced plans to work with interoperability bridges and use its liquidity infrastructure to create deep liquidity pools. Furthermore, decentralized exchanges (DEX) that plan to launch on multiple ZK rollups, such as ZigZag Exchange and ZKSpace (formerly ZKSwap), plan to launch bridges and liquidity aggregation between rollups.


Momentum for ZK rollups is strong. Fiat on-ramps directly to rollups have already launched, wallet support continues to grow, and several Web3 primitives are gradually porting their code to zkEVMs.

Users can get started on experiencing ZK scaling technologies right now, such as by:

  • Using dYdX, Deversifi, ImmutableX, and/or Sorare which all use StarkEx, or deploying contracts and transactions to StarkNet
  • Creating a Loopring L2 wallet and providing liquidity to one of Loopring’s AMM pools (APRs can go as high as 150%+ in certain pools)
  • Sending private transactions at ultra-low fees through
  • Depositing funds into zkSync 1.0 and trade on ZigZag Exchange, which will also soon be live on StarkNet (hint: zkSync has not been subtle about token generation and would likely do an airdrop given the team’s community-first culture)

zkSync Airdrop (Likely) Incoming

Source: @zksync

Using these protocols, users can witness firsthand how frictionless and cheap transactions are through rollups.

By leveraging the security properties of the L1 protocol while providing low-cost, high-throughput transaction capabilities, ZK rollups offer the best of blockchain properties (decentralization and censorship-resistance) with seamless experience that Web2 users expect. Furthermore, with multiple high-caliber teams working on EVM-compatibility, crypto developers should expect to be able to port over their applications relatively effortlessly as well.

Although the future of crypto is likely multi-chain, and even multi-rollup, we expect the emergence of zkEVMs to unlock the next wave of innovation in crypto.

Appendix: Further Resources

Layer Two

  • L2BEAT: Comprehensive overview of the layer two ecosystem
  • L2Fees: Fees across varying layer two scaling solutions
  • Dune: Bridge Away (L1 Ethereum): @eliasimos’s overview of capital locked inside bridges to other L1s and L2s
  • CryptoFees: Transaction fees for various L1s, L2s, and applications

ZK Rollups


The information contained in this post (the “Information”) has been prepared solely for informational purposes, is in summary form, and does not purport to be complete. The Information is not, and is not intended to be, an offer to sell, or a solicitation of an offer to purchase, any securities. The Information does not provide and should not be treated as giving investment advice. The Information does not take into account specific investment objectives, financial situation or the particular needs of any prospective investor. No representation or warranty is made, expressed or implied, with respect to the fairness, correctness, accuracy, reasonableness or completeness of the Information. We do not undertake to update the Information. It should not be regarded by prospective investors as a substitute for the exercise of their own judgment or research. Prospective investors should consult with their own legal, regulatory, tax, business, investment, financial and accounting advisers to the extent that they deem it necessary, and make any investment decisions based upon their own judgment and advice from such advisers as they deem necessary and not upon any view expressed herein.



We help our clients buy and sell cryptocurrency products, earn yield, manage risk and access liquidity.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store