The Scaling Landscape: Current State and Beyond

Inception Capital

Published in

Inception Capital

15 min readFeb 8, 2023

(originally posted 5 Jul, 2022 | Written by Mads Pedersen)

The Big Picture

Types of L2s

Major L2s To Be Aware Of

The Investment Thesis

Other Considerations

Conclusion

The Big Picture

The blockchain trilema states that a chain can only offer two of three core components: scalability, decentralization, security. While Ethereum is both very decentralized and secure, it lacks scalability: Proof-of-Work for distributed consensus has mathematical limits on scalability that are lower than hardware limits. As a result, throughput for Ethereum is limited to the 12–15 TPS range, which is far below the necessary range for mass adoption of the network (Visa can process more than 65,000 TPS). Limited blockspace (supply) coupled with exponential growth in transactions (demand) has resulted in gas fees exceeding $200 for a single transaction for the end user.

Despite the popular belief that “the Merge” with the Beacon Chain into Proof-of-Stake will reduce gas fees, this is not actually the case. This will only alter the way blocks are produced with minor impacts on scalability (TPS) and, by extension, gas fees. The real key to solving this issue is through the implementation of sharding. Sharding aims to increase TPS and lower gas fees for Layer 2s, however, gas fees on Ethereum itself may still be high post-sharding depending on demand.

In response to the scalability issue, Layer 2s and alternative Layer 1s have seen wide adoption.

Layer 2 scaling: Vertical scaling of L1s with verifiable computation with ZKPs (zk-rollups) or a fraud proof (when things break down for optimistic rollups) allows L1 chains to outsource state execution off-chain in highly scalable networks without compromising on security.
Alternative L1s: other L1 chains like Solana, Avalanche, Near, that effectively compete with Ethereum by attracting users by offering lower gas fees and higher TPS.

Both of these solutions have led to fragmentation of users and liquidity. As a result, interoperability has become another large focus to allow for blockchain agnostic, horizontal scaling. Users need a way to transact (and more generally, simply interact) between different Layer 1s, different Layer 2s, from Layer 1s to Layer 2s and vice versa. In a multichain world, how do you fuse siloed ecosystems? Blockchain bridges and multi-chain communication projects have emerged as solutions but are still in their infancy.

Bridges: Infrastructure that enables the transfer of tokens between different chains (very asset focused).
Layer 0s: Multi-chain communication layer that creates the necessary fabric for composability across various L1 to facilitate cross chain communication and horizontal scaling.

Types of L2s

Sidechains

A side chain is a separate blockchain that connects to mainnet but is independent of it. This introduces new trust assumptions since the network has its own consensus mechanism and is not under the control of the L1.

Rollups

Rollups perform transaction execution outside of the main Ethereum chain but post the transaction data to the L1, inheriting the security of Ethereum. This ensures (1) the safety of funds and, (2) permissionless exits from the network as the funds and exit mechanisms are under the control of Ethereum.

There are two types of rollups with different designs.

1) Optimistic Rollups: assume that all transactions are valid and only perform computation in the event of a challenge, involving fraud proofs.

Optimistic rollups do not perform any computation but rather assume that all transactions are valid and aggregators (block producers) are honest when proposing a new state to Mainnet — hence the name “optimistic”. Since relying on good faith would be naive, optimistic rollups rely on a public and permissionless challenging process involving fraud proofs whenever someone notices a fraudulent transaction has occurred. Any network participant can dispute a transaction during the dispute period, where malicious actors eventually have their stake slashed to reward honest participants in the network. Computation is only performed on-chain (Ethereum) when a transaction is disputed, resulting in long wait times for transaction finality (7 days) due to potential fraud challenges and transaction reversions.

2) ZK Rollups: run computation off chain and submit validity proofs (SNARK or STARK) on chain.

ZK rollups bundle large volumes of transactions off chain and generate a cryptographic proof of validity in the form of a SNARK (succinct non-interactive argument of knowledge) or STARK (scalable transparent argument of knowledge) that is posted on chain in the form of a single transaction*. Relayers are responsible for assembling these batches, generating a validity proof, and posting it on chain. This reduces the intensity of compute and data storage on Ethereum as zero knowledge is needed of the entire transaction data. This translates into cheaper transactions, faster finality (compared to optimistic rollups), and lower vulnerability to attacks/fraud. No ZK rollup has decentralized their sequencers yet.

*SNARKS have smaller proof sizes (cheaper) but are not quantum resistant. STARKs are quantum resistant but have large proof sizes (more expensive). The Ethereum Research Foundation has shown their support of STARKs with their $12M grant to Starkware.

Validium

ZK rollups and validiums are relatively similar in how they operate with their main difference being that a validium stores data off chain rather than on chain (external data availability) — making it cheaper but increasing trust assumptions.

Volition

Volitions are also similar to ZK rollups in that they both commit state roots and proofs to Ethereum. The point of difference here is that volitions allow users to choose an alternative data availability solution beyond Ethereum (or the respective L1 it is built on).

Layer 2 Highlights

Rollups are seeing large market wide adoption with $3Bn+ TVL.
Optimistic rollups have seen larger adoption than ZK, which are not EVM compatible yet.
ZK rollups are actively working to achieve EVM compatibility in the near future — zkSync V2 (zkEVM) is live on testnet.
Rollups have started releasing their own tokens instead of using ETH — Optimism’s token, OP, is live and Arbitrum has started its “Odyssey” event which may lead to a token airdrop.

Major L2s To Be Aware Of

Polygon (https://www.polygon.com/)

Polygon offers various scaling solutions and tooling to Etheruem. Their token, MATIC, is traded on Binance, Coinbase, Uniswap, etc. and has a market cap of $3.2B, fully diluted $4B. They last raised $450M in a round led by Sequoia India. They also have a very active community, with 1.5M Twitter followers and 130M+ unique users.

Their largest success so far is Polygon PoS — their EVM compatible Ethereum sidechain — which has attracted 2.7M monthly active users. By offering grants and EVM compatibility, developers are incentivized to easily deploy their applications on Polygon (both new projects and existing live applications on Ethereum Mainnet). A full suite of both familiar and new applications coupled with ~10,000x lower costs per transaction than Ethereum is a strong value proposition for user acquisition. Given that Polygon PoS is a sidechain, there is a trust assumption made because the physical funds and exit mechanisms are not under control of Ethereum.

(Note also: Polygon Hermez, Edge, Avail, Nightfall, Miden, and Zero)

Hermez (live) is a zero-knowledge (zk) rollup optimized for secure, low cost scalability of Ethereum. The zk proof technology allows for the rollup to compute a validity proof of transactions on Hermez, to ensure correctness of transfers and state, which is publicly recorded on the Ethereum blockchain. This way, the physical funds and exit mechanisms are under control of Ethereum.
Edge (live) is a tooling framework that allows users to deploy their own blockchain with customizable features.
Avail (development)
Nightfall (beta) is an optimistic rollup that uses zk technology for transaction privacy aimed at enterprises.
Miden (development) is another zk-rollup that uses zk-STARKs instead of zk-SNARKS as its cryptographic technique. STARKs have the benefit of being quantum resistance, which is not the case for SNARKs.
Zero (development) is another zk-rollup that is fully EVM compatible which relies on recursive proof generation (creating a proof of a batch of proofs) which is faster and less resource-intensive.

Arbitrum One (https://arbitrum.io/)

Arbitrum is the market favorite within the rollup ecosystem. It is the largest optimistic rollup by TVL (Total Value Locked) and has 50.88% (Footprint Analysis) of overall rollup TVL. It does not have a token, so it uses ETH as its native token for transaction fees on its network. The TVL is around $1.9B (L2Beat).

There are clear scalability benefits demonstrated by gas prices:

Optimism (https://www.optimism.io/)

Optimism is an optimistic rollup with security rooted in Ethereum while having a higher transaction throughput and lower transaction fees. Their token, OP, has a market cap of $110M, fully diluted $2.2B. It currently acts as a governance and utility token to pay for transaction fees on its network. The TVL of Optimism is around $713M (L2Beat) but its own token accounts for 16% of this value.

Although Optimism leverages the same technology for its rollup as Arbitrum, there are fundamental differences in design between the two.

Token: most notably, Optimism has a native token while Arbitrum does not
Fraud Proofs: upon a dispute, Optimism uses single-round fraud proofs, meaning Ethereum executes the entire L2 transaction to verify the state root. Arbitrum, on the other hand, uses multi-round fraud proofs, forcing both parties to narrow down the dispute to a single point in a transaction
EVM-dependence: Optimism uses the EVM while Arbitrum has its own AVM (with automatic AVM to EVM translation). This means that Optimism is limited to Solidity while Arbitrum is compatible with all EVM programming languages like Vyper, Solidity, Flint, etc.

zkSync (https://zksync.io/)

zkSync is a ZK rollup developed by Matter Labs live on Mainnet. It has seen growing traction with several larger funnels for liquidity integrating with zkSync, most notably the mobile, non-custodial wallet: Argent. The TVL of zkSync currently sits at $50M (L2Beat) but peaked at $170M in April. Matter Labs boasts a large range of partnerships with major dApps (1inch, Yearn, Connext, etc.) but is largely limited by not being EVM compatible, which is why there are so few projects currently live on its network. However, zkSync V2 (zkEVM) is currently live on testnet and once deployed to mainnet, will drive very high volume to its network due to highly reduced friction for dApp deployment. It does not have its own token and uses ETH as its native token for transaction fees on its network.

Up and coming:

StarkNet (https://starkware.co/starknet/)

StarkNet is a ZK rollup utilizing STARKs instead of SNARKs with its Alpha network live on mainnet since May. Traction is relatively low here with $575K TVL (L2Beat). As of May 25, Starkware (the parent company) raised $100M at an $8B valuation, a 4X markup since its previous round in November. The developer ecosystem on StarkNet is still nascent but is expected to gain traction as the network develops further. Beyond StarkNet, StarkWare builds application specific rollups through StarkEx, with notable applications opting to use this service. However, dYdX, its largest user, recently committed to moving to Cosmos.

Scroll (backed by OP) (https://scroll.io/)

Scroll is in the process of building out an EVM compatible ZK rollup. The next steps for the team are to deploy the rollup on mainnet, begin outsourcing L2 proof, deploy on mainnet and eventually decentralize their sequencers. An excellent team coupled with EVM compatibility and plans for decentralization is very promising for onboarding a vast list of applications and users (currently most ZK rollups use centralized sequencers).

Scroll’s last round was closed at a $350M valuation and they are in the process of raising another round at the moment.

The Investment Thesis

This is just the beginning.

Currently, we are in the early stages of modularity whereby state execution is being abstracted off chain to rollups. While network participants have eagerly moved to utilize these solutions, volume is still low relative to the mainnet itself and the potential of these solutions has only just begun to show itself. See the growth in volume on rollups below:

*NOTE: dYdX is moving to Cosmos

Expect large ZK adoption in the medium to long term.

Optimistic rollups clearly demonstrate greater traction which is largely due to their EVM compatibility. However, this feels more like a short to medium term solution to scalability given the risk of transaction reversion. Once zkEVM is deployed to mainnet and applications can seamlessly deploy to ZK rollups, we will likely see large adoption in this area. Scroll will be a beneficiary of this.

Although both solutions have their respective benefits and drawbacks, there is a clear advantage to having faster finality. The security offered by ZK does not leave room for transactions to be reverted which is appealing to financial verticals within Web 3, namely DeFi, and wider enterprise adoption (where transaction reversion is a large risk). Given that the largest dApp deployed on Arbitrum, GMX, is a perpetual futures exchange, we could see ZK rollups absorbing a lot of the volume from optimistic rollups.

On top of this, ZK cryptography is still very nascent, with wide reaching technological advances still to be discovered. Currently, ZK proofs are computationally intensive, making them expensive and relatively slow to compute. As the technology advances and adoption increases, users and applications will start requiring more specific and expressive statements to be proven which will add to the time and cost of computation. Infrastructure that addresses the bottlenecks and overhead of proof generation and are incorporated into the modular blockchain stack will drive traction by smoothing the process of adoption. A lot of this could very well sit at the hardware level too where just like ASICs accelerate PoW, similar hardware may need to be adopted to ensure that timely proof generation is maintained.

Business development and a long list of partnerships are important — technology alone does not win. With projects operating in the ZK rollup space, customer acquisition and go to market is integral for mass adoption. Currently, the strongest strategy has been through grants, where networks incentivize builders to migrate dApps or start new projects on their respective network. A result of this has been grant farming, where projects choose to build multichain to access as many grants as possible. Newer incentive primitives for builders will start being designed beyond this as attracting talent becomes more competitive — appealing applications drive user acquisition and volume.

Optimizing bandwidth at base layers.

We are currently in the process of transitioning away from monolithic blockchains into a modular blockchain architecture where key functions are outsourced from L1s. As such, we need to begin reimagining blockchains as a modular stack in order to find new investment opportunities that will allow for decentralization, security and scalability at scale. This is all to enable the application layer.

Modularity: A modular blockchain outsources at least one of its responsibilities to another chain.

While rollups considerably improve scalability, they optimize bandwidth for constrained monolithic blockchains. However, modular blockchains can become unconstrained at their base allowing for increased flexibility in use and design. With separation of resources, users and applications are not bound by a single execution environment and can optimize and specialize for specific use cases. Continuous experimentation and development here is necessary to increase the adaptability of previously monolithic chains. The key to survival is adaptation.

Beyond rollups, new solutions are in the process of being built out to further abstract responsibility from Ethereum. Solutions like Celestia are being built out to act as a data availability layer that speaks with both rollups and Ethereum and simplifies the block verification process. Its design means that nodes/validators don’t need to process all the data for the blocks to be valid (data availability sampling), greatly reducing the load on Ethereum for data storage. While optimistic rollups do this too, ZK validity proofs do not since each state transition requires a proof (and proofs are only valid if the transactions in the blocks are processed). Infrastructure that plugs into or builds on top of this nascent modular stack to reduce load on Ethereum are appealing investment opportunities.

Other Considerations

Horizontal Scaling

A natural consequence of various solutions to vertical scaling is siloed ecosystems. We are currently operating in a fragmented multichain world, which will be exacerbated by different rollup ecosystems. Bridges have emerged to connect these different ecosystems in a very asset-centric manner.

Bridges

The Ethereum bridge ecosystem is sitting at over $10B in TVL, peaking at nearly $30B in November 2021.

Bridges play a huge role in allowing for horizontal scaling and interoperability between different chains and will likely continue to do so in the short to medium term. As ecosystems develop further and more users migrate to rollups, bridges will see heightened demand as users will want to move assets between domains.

Similar to blockchains, bridges experience their own trilema of instant guaranteed finality, unified liquidity and native assets.

Instant guaranteed liquidity — the guarantee that funds arrive on the destination chain when the transaction is executed on the source chain
Unified liquidity — the bridge tapping into one pool of liquidity as opposed to multiple
Native assets — receiving the desired native asset being bridged on the destination chain (not synthetic)

Most bridges are designed around lock-mint, where assets on the source chain are locked and kept in escrow and a synthetic asset is minted on the destination chain, or burn-mint, where assets are burned on the source chain and minted on the destination chain. However these designs are subject to the possibility of stealing locked assets held in escrow and unauthorized minting on destination chains, respectively.

These pots of gold are attractive to hackers and have been subject to large exploits. $650M was drained from Axie Infinity’s Ronin bridge, around $325M was drained from Wormhole, $80M was stolen from Qubit Bridge, and the list goes on. As such, investing in robust cross chain liquidity networks is of high importance.

Connext (backed by OP)

Connext is a bridge designed around swaps, where users swap existing assets (no minting). $1.2B worth of assets has been bridged through Connext to date, with $45M available liquidity, 650K total transactions and 44 live routers. With strong tailwinds in traction and robust engineering, Connext proving itself to be a strong player in the bridging space. On top of this, it is positioning itself to become a more general interoperability protocol beyond asset-centric interoperability, where its bridge is just the beginning of the story.

L0s

A wider solution to fragmented chains and horizontal scalability are L0s that act as generic architecture support that sit below L1s to enhance scalability (optimized topology, block compression) and allow for cross chain communication.

L0s typically allow for faster communication between nodes (optimized latency) instead of reliance on P2P messaging and compression of transaction data for faster block propagation. However they can also be built to form a fabric for cross chain messaging beyond the focus of assets. With this technology, users will eventually be able to engage with any chain without leaving their source chain.

Case study: Project XYZ mints an NFT collection on chain A but builds a game around this collection on chain B since gas fees are too high on A. What now?

A L0 would be able to integrate with this NFT collection and game and allow NFT holders on chain A to play the game on chain B without ever leaving A.

Although this a simplified, more trivial example, the technology has far reaching implications for wider interoperability that bridges cannot capture. In fact, bridges can be built on top of this cross chain communication architecture (see Stargate on LayerZero). As such, given the demand for cross chain interaction (shown by bridges) and vast list of different L1s and L2s, it is inevitable that the demand for L0s will only increase to prevent the isolation of chains.

Conclusion

Overall, it is clear that the current scaling landscape is still nascent and there is room for growth in the infrastructure that supports it. Some of the most important aspects of this are vertical scaling through rollups and horizontal scaling through bridging/cross chain communication. At OP Crypto, we are committed to advancing the modular infrastructure stack to alleviate the growing pains of monolithic chains, enabling the sustainable onboarding of hundreds of millions of users onto the application layer. This will translate into both direct profit for the fund and indirect profit for the entire Web 3 ecosystem that sits on top of (and below) this technology.

ABOUT OP CRYPTO

OP Crypto is a leading high conviction, early-stage venture capital firm in the crypto and blockchain industry, specializing in pre-seed and seed stage investments. The fund successfully raised $50M in September 2021 and has since invested in over 30 projects, including companies such as Scroll Tech, Snackclub, Merit Circle, Omni, and Fyde.

With the support of prominent investors like Bill Ackman and Alan Howard, as well as institutions like Galaxy Digital, Huobi, and DCG, OP Crypto has access to a global network of venture funds, scouts, and ecosystem partners to source the most competitive deals in the market.

With a core team based in the United States and strong ties to the APAC region, the fund serves as a bridge between East and West. Additionally, a dedicated portfolio team provides post-investment support to founders in areas such as marketing, tokenomics, and networking.

Learn more about OP Crypto at: Website