An Investor’s Perspective on Rollups
The crypto community is no stranger to the issue of scalability on Ethereum. As the DeFi and NFT market grew, people were placing more transactions at any given time, placing a huge strain on the Ethereum network. The computational demand caused the network to become congested, thus making it more costly to execute transactions. The figures below shows the mean gas used (i.e., paid) across all transactions.
Today, the high cost of executing even the simplest transaction has caused many people to leave Ethereum for other blockchains or shy away from crypto altogether. Gas has become the Great Gentrifier, pricing out the everyday user, discouraging experimentation and devolving DeFi into basic yield farming. Gas is the Helen of Troy, subjecting peaceful NFT communities into dreaded Gas Wars.
The Future is Rollups
For years, the crypto community has been waiting for the advent of Ethereum 2.0 (Eth 2.0), the Ethereum Foundation’s promise of shifting the blockchain away from Proof-of-Work and into a more scalable and less costly Proof-of-Stake model. Ethereum has hit a number of significant milestones in the Eth 2.0 roadmap ever since the launch of the Beacon Chain on December 1, 2020. Meanwhile, Ethereum has been moving forward with the rest of its phases in parallel:
- The Surge — increasing the scalability of rollups through sharding
- The Verge — implementing stateless clients to enable lightweight transaction verification
- The Purge — eradicating old unnecessary data to reduce technical burdens
- The Splurge — adding non-critical enhancements to Ethereum’s design
As we enter the Surge, I’ve been noticing a lot more projects building with rollups than I had just a few months ago (For those who are unfamiliar with rollups, I recommend reading Vitalik’s Incomplete Guide to Rollups). Of course this technology is not new, and there are tons of well-known companies in this space, but lately I’ve come across a lot more pre-seed startups building with rollups as Eth 2.0 begins the rollup-centric stage of its development.
Why are rollups so important, and why has the Ethereum Foundation dedicated an entire phase of the Eth 2.0 roadmap to rollups? Rollups are the key to helping Ethereum achieve scalability. Prior to 2020, the goal of the Eth 2.0 roadmap was to have Ethereum layer 1 (L1) execute all transactions by spreading the network load across 64 separate shards. Since then, Ethereum’s overall strategy has changed. Rather than acting as the execution layer for all end user transactions, the L1 will become a settlement layer for rollups, serving as the security and decentralized consensus layer for a rollup-centric layer 2 (L2). Ethereum will therefore transition into a completely L2 ecosystem as all end user transactions occur on rollups. In fact, the Ethereum Foundation is no longer even using the term “Ethereum 2.0,” referring to Ethereum as a “consensus layer” instead.
The figure below depicts the structural makeup of Ethereum once its L1 becomes a settlement layer. Block production occurs on rollups, and block validation occurs in Ethereum’s 64 shards. The Ethereum blockchain can handle 15 to 20 transactions per second (TPS). With the proper compression technology, rollups can handle 1,500 to 4,000 TPS. Only with sharding can rollups reach 100,000 TPS, which is the desired level of bandwidth that Ethereum will strive to achieve during the Surge.
The increase in projects building with rollup technology makes sense within this context. Ethereum has made significant progress during the Merge, and it is now preparing to become the settlement layer for rollups. Meanwhile, companies are racing to create the most optimal rollup, while others seek to build the protocols that live on top of rollups.
A Rollup-Centric Ecosystem
As companies continue to make advancements in rollup technology, two possible scenarios arise:
- One rollup dominates the ecosystem because it achieves meaningful node implementation, processing at least 10,000 TPS, and Ethereum stores the rollup’s on-chain data between its 64 shards.
- No single rollup is able to process more than a few hundred TPS, and a multi-rollup ecosystem is born. Numerous rollups are processing transactions at any given time as Ethereum’s base layer provides data availability and shared security. Users are constantly switching between various rollups using cross-rollup bridges.
I predict we will end up in the second scenario. Rather than having one rollup achieve hegemony, rollups will gravitate towards heterogeneous scaling, which means each rollup will have different tradeoffs of the blockchain trilemma depending on the needs of various applications. For example, one dApp may need to build on top of a rollup that can handle fast withdrawals, while another dApp may need a rollup that provides faster finality time or better interoperability.
We are already witnessing the beginning of a multi-rollup ecosystem. A vast majority of the projects depicted in the graphic below have integrated some form of rollup technology, and I can already name a few projects that are not included in this list.
Looking Forward
Like in any ecosystem, progress is made through trial and error. Rollups are still a nascent technology; people are trying to figure out its limits, how to make it more scalable, and when best to use it. As an investor, these are some of the main themes I will consider when I look at projects building with rollups.
Rollups may not be the best solution. At the moment, rollups are too expensive for many users. Since rollups must post call data to Ethereum, their fees are inextricably tied to Ethereum fees. Though there are many companies aiming to make rollups more efficient, there are L1s and other forms of L2s in the market that already provide a more cost-effective scaling solution. As shown in the chart below, rollups like Optimism, Arbitrum, ZKSync, and Loopring generate higher fees than alternative L1s like Solana and Avalanche.
Whenever I come across a new protocol, I will evaluate whether it made sense for it to be built on top of a rollup, sidechain, or alternative L1. Each scaling solution has its strengths and weaknesses. Haseeb Qureshi from Dragonfly Capital provides a compelling analysis of how rollups match up against other scaling solutions.
Cross-Rollup Bridges. Once users have completely migrated onto rollups, there needs to be infrastructure that allows people to bridge from one rollup to another. A great example of an interoperability protocol is zkLink, which not only aims to connect L2s with other L2s, but also L2s with EVM and non-EVM L1s.
I am interested to see how the interoperability landscape evolves as asset-specific, chain-specific, app-specific, and generalized cross-chain projects all compete for liquidity. Bridge protocols will either need to maximize the liquidity they control by connecting as many chains as possible or become specialized in order to root out competitors. I believe we will mainly see scenario 1 play out in the beginning, but we will see scenario 2 play out over time.
L3s are coming. Another consequence of shifting user activity completely onto Ethereum’s L2 is the need for an application-specific L3. L3s offer more customizability to developers, allowing them to not only adjust the performance and costs of applications, but also to add special features. These features may include hyperscalability as the compression benefit of L3 proofs multiply that of L2 proofs. L3s can also serve as privacy-preserving protocols on top of a public L2. L3s can even act as a canary network for developers to test out their protocols before releasing them to the general public on an L2. StarkWare discusses the numerous potential benefits of L3s in more detail.
Overall, I am most excited to learn from founders building new protocols and dApps made possible by the upcoming advancements in rollup scaling technology. The collective imagination of the crypto community has been constricted by the limits of technology for too long, and I’m optimistic that rollups will bring us one step closer to bridging our ideations with reality.
