Ethereum: the L2 war can begin

After years of development, there are now nearly a dozen secondary layer protocols on Ethereum, and each wants to establish itself as the benchmark L2. Who has the best weapons? We investigated. 🔍

Eight years after its launch, there is no longer any room for doubt: Ethereum has truly established itself as the reference blockchain for the deployment of smart contracts and decentralised applications (dApps).

Where nothing is certain, however, is at the level of second-layer blockchains known as "layers 2", as opposed to layers 1 such as Ethereum.

Layers 2 have emerged gradually to solve Ethereum's development problems. During peaks in activity, Layer 1 is unable to handle all the transaction traffic, which congests the network, causing transaction fees to skyrocket and delays in confirming a transaction. In short, Ethereum doesn't work very well for scaling up! 😔

His co-creator, Vitalik Buterin, had already mentioned this problem several times, which he sums up as the "blockchain trilemma".

According to him, a protocol cannot be both:

  • "scalable", i.e. capable of going to scale
  • "secure"
  • "decentralised"

One of these three elements must always be sacrificed for the benefit of the other two. In the case of Ethereum, it's "scalability" which, no pun intended, has borne the brunt of the billiard shot.

The downside of this choice is that the Ethereum blockchain is therefore doomed, at least in the medium term, to be rather slow and expensive. It is precisely to solve this problem that a myriad of second-layer protocols, our famous "layers 2", have been created.

Their aim is to make Ethereum faster and cheaper. 🚀

A layer 2 is an additional layer that works in parallel with the main blockchain, Ethereum. These solutions allow part of the processing of operations (transactions, smart contract executions, etc.) to be carried out outside the latter, thus reducing the load on the main network and the associated costs.

Polygon has attracted the big brands so far

There are different L2s, including sidechains and rollups. Sidechains are independent blockchains that rely on Ethereum and transfer part of the transactions to this secondary blockchain.

The main shortcoming of these sidechains is that they have to provide their own security. 🥵

In this category, Polygon PoS is the most popular project, particularly with brands. In less than three years, the sidechain has attracted several large traditional companies such as Reddit, Meta, Starbucks and Nike to NFTs projects. There are dozens more.

"Polygon is particularly effective on business development," confirms Stanislas Barthélémi, crypto expert for KPMG. "However, they have lost traction in recent months in favour of more recent and innovative technologies such as rollups."

"Rollups"

As with sidechains, rollups make it possible to manage a large number of transactions by combining them into a single transaction on the main Ethereum blockchain. But unlike sidechains, rollups rely on Ethereum's security and therefore don't have to manage it itself. 😍

This difference is a major advantage, as Ethereum is one of the most secure protocols in the sector. Using its security means ensuring very strong resilience.

There are dozens of rollups on the market, and new ones arrive very regularly. This week, US giant ConsenSys, the company run by one of Ethereum's co-creators, Joe Lubin, launched its own rollup: Linea. All the details on the project are here.

There's also Base, which is being developed by exchange platform Coinbase. More are expected to follow. 📅

More broadly speaking, there are two main families of rollups: optimistic rollups and ZK rollups.

L21

(Value immobilised in the various L2s since 2019 - L2Beat)

Optimistic rollups vs ZK rollups

The difference between these two types of protocol is quite simple. Optimistic rollups, as the name suggests, are "optimistic", meaning that all transactions are presumed to be correct and therefore integrated into Ethereum.

To ensure that the system is viable and that "bad" transactions are not processed, there is a mechanism that allows transactions to be challenged. This mechanism is open for a period of a few days. ⏳

The big strength of optimistic rollups is that they are very practical. They have the same virtual machine as Ethereum: Web3 developers can therefore launch projects on them very easily.

The weak point, because no project is perfect, is precisely this delay in contestation, because you have to wait until it has elapsed for the transactions to be officially integrated into Ethereum. "It's not very comfortable when you want to withdraw your funds quickly," confirms Jérôme de Tychey, president of Ethereum France, which is organising the Ethereum Community Conference (EthCC) next week in Paris. 🇫🇷

The EthCC is one of the world's biggest conferences for the Ethereum ecosystem. Many events are planned around L2s.

The best-known optimistic rollup projects are Arbitrum and Optimism, which account for 65% and 22% of total L2 activity respectively, according to data from L2Beat. This success is mainly due to the fact that these projects were launched earlier, in 2021.

The fact remains that optimistic rollups are not immune to competition, which comes precisely from the second family of rollups: ZK rollups.

The special feature of ZK rollups is that they enable transactions to be validated without revealing any confidential information (Zero Knowledge Proofs). How do they do this? Thanks to cryptographic techniques.

The (very) big advantage of this system is that ZK rollups work without the system of time limits for disputes. It's virtually instantaneous. ⚡

Much more recent, ZK rollups don't yet have the same traction as optimistics rollups. The best known of the ZK rollups is zkSync (Era). The other that is emerging is Polygon zkEVM (Polygon's rollup proposition).

"The difficulty with ZK rollups is having a tool that uses Zero Knowledge Proofs that is compatible with Ethereum," explains Jérôme de Tychey. "This system is probably the most interesting and the most innovative, but its development is very complex and expensive", he believes.

To overcome this problem, projects such as StarkNet or Loopring have chosen slightly more pragmatic approaches. In the case of StarkNet, this involves using Zero Knowledge Proofs with machines other than the Ethereum machine. The consequence is that project development is a little more complicated for developers 😅.

L22

(The market share of the various L2 projects - L2Beat)

Emulation between the various projects

Paradoxically, the L2 war benefits the entire ecosystem. "The solutions complement each other quite well," notes Jérôme de Tychey.

"The projects work to achieve a common goal. It's a bit like with vaccines: there are several approaches, but the collective goal is to come up with a solution against a virus", stresses the man who is also the founder of the gaming Web3 studio Cometh.

This observation is shared by Stanislas Barthélémi of KPMG: "The head-on opposition between optimistic and ZK rollups has lost its relevance. We are seeing more and more projects that borrow from both universes". 🤝

From this point of view, the discussions within the Optimism community are an excellent revealer: some are arguing for the optimistic rollup to incorporate the Zero Knowledge Proofs.

Some projects like Linea are even arguing for the convergence of the solutions on the grounds that their similarity will make it easier for developers to switch from one to the other. "The fewer fundamental differences there are between the L2s, the more the ecosystem will develop in a decentralised way", stresses, Nicolas Liochon, in charge of research and development at ConsenSys and Linea.

"I nevertheless think that we will probably be 30% faster than the others", he breathes, highlighting the unavoidable race for numbers in which each project is engaged. ⚔

Communication battle

Because while some are playing it "collective", the reality is that each project - Linea is not alone - wants to boast the best performance.

Starkware, the Israeli start-up developing the StarkNet network, last week presented its new "Quantum Leap" update, which is said to be - potentially - capable of handling "several hundred transactions per second".

According to data recently recorded on its test network, this new iteration of StarkNet was capable of handling 90 transactions per second. A feat, when you consider that competitors rarely exceed ten or so transactions per second - in real-world conditions. 😲

The leap forward made by Quantum Leap is excellent news for the sector and demonstrates its capacity for innovation. However, some people are questioning the point of such competition on this kind of performance.

"Ease of development and interoperability are certainly more important arguments," Ethereum co-creator Joe Lubin (we've got a big interview in store for you next week 🔥 ) tells The Big Whale.

Many are criticising StarkNet for opting for its own Cairo programming language (which differs from Ethereum's Solidity, ed. note), reputed to be rather difficult to get to grips with.

"We provide access to all the most important features so that crypto reaches mass adoption," counters Louis Guthmann, in charge of ecosystem development at StarkWare.

So a battle of the numbers is well and truly underway, even if it makes little sense due to relative demand.

This press release-driven rivalry is inevitably reminiscent of the time when "Ethereum Killers" such as Solana or Elrond, blithely banking on the - mostly theoretical - performance of their networks.

"Although StarkNet is a really interesting proposition, it's only when there are large-scale users that we'll be able to really measure its performance," judges Stanislas Barthélémi. 💡

Before StarkNet, the start-up Starkware (valued at $8 billion) had made a name for itself by developing a solution that enabled NFT giant Sorare to offer virtually non-existent transaction fees on its fantasy football game.

This product, called StarkEx and tailored to the needs of projects, recently passed the $1 trillion transaction threshold. StarkNet is its "open source" offshoot.

The question of project tokens

In addition to the theoretical performance communicated, the other data to watch out for when measuring the attractiveness of an L2 is the presence of a governance token.

Some projects have already issued one, following the example of Arbitrum and Optimism. The effect has been fairly immediate, with an influx of users, delighted to be able to earn money quite easily. With their tokens, the protocols are dousing the first users in order to stimulate activity on their network. 🤑

"This kind of strategy can be effective at first, but the problem is that you also attract a lot of people who are only there to collect tokens and who will leave the adventure as soon as the tap turns off," warns Stanislas Barthélémi.

"Only organic growth, i.e. growth from users who come because of the quality of the technology on offer, ensures the viability of a protocol", he assures.

Among the projects that have not yet issued tokens are zkSync and StarkNet in particular. However, the latter should not be long in launching.

🔴 According to our information, StarkNet's should arrive before the end of 2023. As for Linea, the latest entrant, the subject is still far from settled.

The designers believe that there would be little point in having a governance token. And with good reason: there is currently only one operator on the network, and it is managed by... ConsenSys. "When we are sufficiently decentralised, with several sequencers, then we will eventually be able to think about it", believes Nicolas Liochon.

This strategy is close to that of Base, the L2 developed by the American exchange platform Coinbase. This is based on technology developed by Optimism.

"A network that is not decentralised has no reason to issue a governance token," insists Nicolas Liochon. This only serves to enrich project sponsors in the short term, but this type of strategy proves disastrous in the long term."

Is there at least one L2 project that is fairly decentralised to date? 🤔

"At the moment, no L2 is really decentralised because there is still a lot of research and development to be done before we can consider decentralising governance," says Stanislas Barthélémi. "In the end, it is more likely that the projects that are the most cautious about governance will be the best placed to deploy their solutions," he concludes.

Everything that matters in Web3. Each week.
25€/month
Try insider for free, for 30 days.
Subscribe
All that matters in crypto.
Deciphering, insights, Data. Access the best of the ecosystem.
Subscribre
In this article
No items found.
Read next
No items found.
In this category