Rollup rollup: scaling blockchain to one billion users

Crypto and especially Decentralized finance (DeFi) promised to democratize finance, the same way the internet made content accessible to everyone.

The first commercial Bitcoin transaction back in 2010 was to buy two Papa John’s pizzas (as an Italian I have to admit, that hurt), with the moment heralded as the first step towards widespread everyday crypto payments.

More than a decade on that’s still nowhere in sight. Bitcoin and Ethereum support 7 and 15 transactions per second (TPS) respectively compared to the 5000 and 24000 TPS of Mastercard and Visa.

Supply and demand dictates that transaction costs surge with heavy use. With particular load heaped on Ethereum due to the explosion in popularity of DeFi, NFTs and crypto gaming which all run on the protocol, average transaction (gas) fees have increased from $2 a year ago, to $50 in November 2021.

If Ethereum wants to power the vision for the web 3.0 and metaverse (check out our overview of NFTs and crypto gaming) it needs to enormously scale its capacity.

The quick way to achieve scalability would be to sacrifice decentralization or security (AKA the blockchain trilemma). The Ethereum community has instead chosen to execute a wide scalability roadmap towards Ethereum 2.0. Rollups in particular are poised to be the key scalability for Ethereum in the foreseeable future and have already reached almost $6B in Total Locked Value (TLV), up 8x in the last months.

But what is a rollup, what are modular blockchains, and why should we care about the blockchain trilemma?


Why is it so hard to scale? To understand it we have to go back to blockchain core principles.

Subscribe to Dealroom’s Fintech newsletter, for weekly insights on finance-focused startups and investment:

The blockchain trilemma

The blockchain trilemma means one blockchain cannot be optimized at the same time for scalability, security and decentralization.

As mentioned, the explosion of popularity of DeFi this year has brought too much traffic on the chains like Ethereum making fees skyrocket. As a consequence decentralized finance is becoming dominated by whales (big traders), with average transaction sizes in protocols ranging from 10k to 1M.

This has opened the opportunity for competing Layer1s, the so-called “Ethereum killers”, to attract users away and new users to their network with much lower fees. The basic issue is that to have a still secure but much scalable blockchain chain they had to give up decentralization. They grew fast but still Ethereum’s total locked value (TLV) reached $178B, almost 10x than nearest chains Binance and Solana.

How can Ethereum become scalable while maintaining decentralization? Here comes modularity.

Breaking the blockchain trilemma: modular blockchain

As of today, all blockchains are monolithic. This means that they host execution (upgrading the blockchain state), consensus (providing security and defining the canonical truth) and store the data, all on the Layer1.

Modular blockchains instead can separate these functions. In particular, the vision for Ethereum is to move execution away from the Layer1 to Layer2 solutions such as rollups. For more info check out the Bankless Ultra Scalable Ethereum thesis.
Comparison between monolithic and modular blockchains

Scaling solutions: a spectrum from side chains to plasma, channels and rollups

Several scaling solutions have been investigated through the years. Without getting into too many details, solutions like State Channels and Plasma are “full” layer2 since they move both execution and data availability off-chain. This brings a limited safe use for generalized applications. They are now considered deprecated.

Sidechains solutions instead move execution off-chain and use their own consensus mechanism. The most successful example is Polygon Matic which has almost $5B in TLV. This is a sidechain connected to Ethereum which uses PoS and therefore requires validators to stake its native Token (Matic) to secure the network. Polygon is also rumoured to be raising 50-150M from Sequoia Capital India and Steadview Capital through token sales.

But Polygon is much more, it is a portfolio of solutions or as they describe it the “Ethereum’s Internet of Blockchains”, fully dedicated to scaling Ethereum. In particular, Polygon sees Zero-knowledge rollups as the ultimate holy grail of scalability. It already has 3 public solutions (Nightfall, Hermes, Maiden) and a fourth one to still disclose. It is also committing $1B from its treasury for the development of zk-based solutions and it has just acquired ZK-rollup Mir protocol for $400M!

Rollups move computation (and state storage) off-chain but keep some data per transaction on-chain. Rollups are expected to be a cornerstone of Ethereum scaling in the short and medium-term future (and possibly long-term as well). Rollups are still an early-stage technology, and development is continuing rapidly, but they already work, some have been running for months and have amassed billions of values and hundreds of applications.

They also raised almost $400M from investors, led by Offchain Labs (Optimism), Starkware, Immutable, Matter Labs and Optimism. And this is just a small part of the money since as we mention only Polygon is investing $1B in zk-rollups and even newcomers like Optimistic rollups Metis have raised a $100M fund to support its project ecosystem.

Optimistic and ZK rollups

Rollups can mostly be divided into two categories: optimistic and zero-knowledge (ZK).
Optimistic and Zk rollups basic principles

Optimistic rollups assume that the transactions are valid by default and only run the computation, through a fraud proof, in the event of a challenge. Basically, if you submit a transaction you also lock up a part of your ETH, so if your transaction validity gets challenged and your fraud is discovered you lose part of it. This process can take up to a week.
ZK rollups instead run computation off-chain and submit a validity proof to the chain.
ZK rollups bundle hundreds of transfers off-chain and generate a cryptographic proof (SNARK). This is known as a validity proof and is posted on layer 1.

Which are then the pros and cons of these two architectures?

Basically, optimistic rollups are a bit easier and have the short term advantage of EVM (Ethereum Virtual Machine) compatibility, which means Ethereum smart contracts can be deployed directly on these chains. That’s why DeFi based platforms are using them.
But their week-long time to confirm the transaction (finality time) prohibits a seamless interaction with L1. Applications like NFTs, crypto gaming are not compatible with this and even many DeFi applications are limited. That is why in the medium to long term ZK rollups are expected to win out in all use cases.

For a more detailed comparison see the table below.
Comparison between optimistic and zk rollups

Rollups should be able to bring Ethereum to ~1.000-10.000 TPS, further scaling combining rollups with sharding has the potential to bring to 100.000+ TPS. For a more in-depth explanation, you can refer to Ethereum 2.0 rollups guide. We will also not discuss here sharding either. Here is an in-depth overview of the Ethereum roadmap by Vitalik.

Rollups adoption and main players 

Layer2 scaling solutions have now reached almost $6B in TLV, led by general-purpose optimistic rollups (such as Arbitrum, Boba Network and Optimism) and exchange and payment focused ZK rollups (such as dYdX, Loopring, ZKSwap). General-purpose zk rollups are just now going live, two of the leading teams competing in the space are Matter Labs (with zkSync 1 and 2) and Starkware (with StarkEx and StarkNet).

Beyond the technical capabilities and developments, these networks will compete on user onboarding. Layer2 solutions are unfamiliar for users, still few applications are deployed on layer2 compared to layer 1, wallets are less available, merchants acceptance is lacking. All these independent projects are competing with each other to scale Ethereum, while also competing with alternative Layer1s to retain and attract users to Ethereum. They are in a sense the business development arm of Ethereum.

Launching funds to boost and subsidize the ecosystem is the strategy being adopted by Layer 1 competitors. Will layer2 scalability start to do the same? Optimistic rollup Metis has launched a $100M fund “Genesis” for this. An experiment to watch.

The importance of network effects and the risk of a fragmented future

Ethereum value heavily resides, like every network, in its network effects and the virtuous cycle between developers and user adoption. If we assume there will not be a single scaling solution that “wins” the scalability wars, this means we will have several Layer2 scalability solutions bringing fragmentation. Two of the main issues would be composability and liquidity fragmentation. Composability means protocols can interact with one another, building on top of other protocols and combining functionalities seamlessly. This might be limited if Layer2s cannot communicate easily without the need to pass on the Layer1 with its high expenses. Solutions like cross-chain cross-chain rollups or interoperable layers like Polygon could efficiently move assets and data from one rollup into another. But it will be a long journey to get all Layer2s solutions to agree to a standard framework.

A similar problem is liquidity fragmentation. Liquidity is the bandwidth of the crypto and web3 world and its fragmentation among the different Layer2 would severely hinder the capacity of protocols. Several strategies are being adopted also here. For instance, DeFi pooling allows users on Layer2 solutions to interact with liquidity on Layer1 or exchanges (StarkEx 3.0 has launched a similar solution).Explore the layer2 scaling solutions for Ethereum on Dealroom

Will Ethereum manage to pull off its ambitious scaling strategy, or will it lose part of its dominance to other networks?

Subscribe to Dealroom’s Fintech newsletter, for weekly insights on finance-focused startups and investment: