Speech: Vitalik Buterin, Founder of Ethereum

Source: Chain News

On October 26, the 7th Wanxiang Blockchain Global Summit was held at the W Hotel in Shanghai. Vitalik Buterin, the founder of Ethereum and Chief Scientist at Wanxiang Blockchain Lab, shared his views on the current hot topic of Layer 2 scaling through an online live broadcast. Below is the full text of the speech:

Recently at the Wanxiang Summit, I have mentioned multiple times, and at other summits as well, the topic of scaling, scaling is one of the biggest challenges facing the blockchain field. This includes the biggest challenges we face today, such as the fact that while blockchain is very effective in decentralization and ensuring security, it can only handle a very small transaction volume, meaning its TPS is very limited. A few years ago, this state of blockchain did not pose a significant problem for us because there were not many blockchain users. However, recently we have seen many people wanting to use blockchain, and this number is growing daily. Moreover, people want to do more and more things on the blockchain, and the number of such people is also increasing. Now various applications have emerged, such as DeFi, Gamefi, etc., with many application scenarios that people want to realize. This means that blockchain must achieve scaling to provide the space needed for all applications.

There is a simple way to achieve blockchain scaling and a difficult way. The simple way is to increase parameters such as block size and gas limits, allowing us to handle a significantly larger volume of transactions. However, the problem with this approach is that it seems simple but sacrifices decentralization. Achieving such scaling by creating a massive system means that each individual node in this massive system needs to handle a large amount of data. Such a system is likely to lose decentralization because very few people will operate such large nodes.

Another method to achieve blockchain scaling, which is also more sustainable, is Layer 2 technology. Layer 2 technology is diverse and complex, including some lightning networks, as well as some applications that have already used state channels, and other solutions used by various applications.

The most common technology we see is Rollup. The concept behind this technology exists in many different forms; it began in 2014 and took shape in 2019. In 2019, I published a blog post, as shown in this PPT, where I wrote about a technology roadmap centered on Rollup. I mentioned that we need scaling, but for Ethereum, achieving native scaling will take a long time, specifically executing sharding, where transactions can be processed within each shard, and achieving this requires overcoming many technical challenges and will take a long time. Therefore, we must wait a long time to truly realize Ethereum's native execution of sharded scaling; the solution we have now is Rollup. We have already seen different versions of Rollup projects emerging in the market, and we know that Rollup is useful and universal. Rollup can significantly enhance scalability, at least by a factor of 100 so far. Data sharding is a relatively simple form of sharding; it does not create a powerful shard capable of processing transactions; rather, we only create shards that can store and validate data, which is a simple form of sharding.

Through such sharding, we can improve the scalability of Rollup, potentially increasing it by another 100 times. As you know, Rollup can provide us with scaling, and we are well aware of this. For Ethereum, Rollup is the only viable scaling solution.

The difference between last year and this year is that when we used Rollup for scaling last year, most of the projects were merely theoretical. Last year, Loopring existed, but other projects did not. However, this year, many Rollup projects have launched their mainnets, and their transaction fees are significantly lower compared to Ethereum's transaction fees. I believe these numbers, meaning their transaction fees, will further decrease, and I have seen many optimizations in this area. Regarding how to use Rollup to reduce the amount of data that needs to be stored on-chain, this method can further lower the gas fees required for Rollup transactions. For example, Loopring has used many optimization solutions, making it at least 30 times cheaper than Ethereum.

There are also Arbitrum One and Optimism, which are two Rollup solutions that currently have not implemented any optimizations, but they are working hard. Once they incorporate optimization solutions, these two Rollup solutions will also become cheaper. Therefore, with such Rollup solutions, transaction fees can drop significantly. They can be much lower than last year's transaction fees, and if sharding can be realized, it can further reduce transaction costs.

We can now see a variety of Rollup technologies on the market, and Rollup itself has many technical components. Today, there are two main branches of Rollup: one is Optimistic Rollup, which uses a security scheme, and the other is ZK Rollup, which uses zero-knowledge proofs. Both Optimistic Rollup and ZK Rollup have their respective advantages. For Optimistic Rollup, its technology is simpler, making it relatively easier to build Optimistic Rollup, especially for general-purpose applications, such as general-purpose applications like Ethereum smart contracts. ZK Rollup technology is more complex, but its benefits include higher security and support for fast exits, unlike Optimistic Rollup, which requires a week to exit. Technically, ZK Rollup technology is not yet fully mature, but I believe that in a few years, ZK Rollup will support a complete version of the Ethereum Virtual Machine, allowing everyone to use ZK Rollup technology. In the short term, Optimistic Rollup is a feasible choice, while in the long term, I prefer ZK Rollup. Once we have data sharding, ZK Rollup will also have other applications. I am also considering some enterprise-level applications; once ZK Rollup matures, enterprise-level applications will become possible. There is no need to use a separate chain; you can use ZK Rollup, which connects to Ethereum, keeping costs low while also benefiting from the interconnectivity with the Ethereum ecosystem and the advantages provided by the Ethereum public chain.

This technology has made significant progress, with many projects gradually migrating their applications to Rollup. Many projects have already begun to think about how to build their projects within Rollup. However, there is still much work to be done, such as applications and infrastructure, and there are many things to accomplish.

If you are an application developer, you must actively seek how to migrate your application to Rollup as quickly and safely as possible. For non-financial applications, migrating to Rollup is even more important because it can significantly reduce costs. After all, reducing costs is very important for non-financial applications, as their users are less willing to pay high transaction fees. In addition, many applications are a complete system, meaning that if users use this application, they are more likely to want to interact with other applications. For example, in the game "Dark Forest," which is a completely trustless game based on Ethereum, this game is a great example. It is currently running on the testnet and can be migrated to Rollup in the future; NFTs can also be migrated to Rollup. This means that infrastructure can be migrated to Rollup, and financial services can also be migrated to Rollup, such as applications that use cryptocurrencies for payments, which can also be migrated to Rollup, and DAOs can soon be migrated to Rollup as well.

For different applications, the entire migration process varies greatly. Each application must consider whether to fully migrate to a Rollup, support all Rollups, or continue to reside on Ethereum while establishing cross-chain bridges with Rollup. Of course, there are many paths to achieve this, and each option needs to be carefully considered. Next, I will discuss the core part of the entire ecosystem—infrastructure.

For users, infrastructure is key. ENS represents the Ethereum Name Service, which is a decentralized domain name system based on Ethereum. ENS is a very important infrastructure, and many Ethereum users use ENS; for example, I have an ENS domain name, and many Ethereum users also have ENS domain names. My own blog also has an ENS domain name. If you visit my blog, you can see it. In the future, I expect more and more people to use ENS. The current problem facing ENS is that all functions are implemented on the mainnet, which leads to very high transaction fees, and this is precisely the challenge we face in infrastructure. As we migrate ENS to Rollup, this issue is expected to be resolved.

The second piece of infrastructure is light clients. If we want the Ethereum system to achieve true decentralization, we need secure light clients that can not only access Ethereum's mainnet but also access cross-chain bridges and Rollups. This means it should be easy to move from one Rollup to another, facilitating cross-chain interactions without long wait times.

EIP-3668 is a proposal that allows ENS and other applications to migrate to Rollup. It is a general proposal that also allows Layer 1 smart contracts to specify where to obtain application or Layer 2 state information and to validate that information, meaning that a light client can validate Layer 2 data without requiring verification. Additionally, you can register an ENS domain name and update it on Layer 2. Regardless of which Layer 2 it exists on, users can see this information and use the same software to view information across various Layer 2s. This prospect is exciting.

Light clients allow a node to validate the blockchain without needing to run a full client or validate every transaction. Light clients are crucial for achieving decentralization in the blockchain ecosystem. What we are currently doing is upgrading through PoS and supporting light clients, ensuring that the light client system is equally robust on both PoS and PoW. It has an algorithm that allows light clients to validate blocks. These blocks are signed by a committee, enabling secure clients to access the blockchain without trusting a centralized server. However, the light client system itself also needs further expansion to migrate into the Rollup system. This is very important; we still need to maintain decentralization, and we are continuously working towards applying this upgrade to all Layer 2 protocols.

Let me conclude.

Layer 2 is the future of Ethereum scaling. Using Layer 2 to achieve Ethereum scaling is currently the safest solution. For Ethereum's scaling, it is crucial to preserve the network's decentralization while achieving functionality, as decentralization is the core of blockchain.

Currently, Ethereum Virtual Machine-compatible Rollups have emerged, providing strong scalability while ensuring performance. At the same time, it maintains the current user experience and developer experience. Therefore, developing for Rollup is akin to developing on Ethereum, with only minor changes. For example, the method of block generation and packaging is different, but its virtual machine is the same, and compatibility with Ethereum's virtual machine means preserving all of Ethereum's performance while maintaining the very important attribute of Ethereum's decentralization.

Meanwhile, we also need to continue building infrastructure, including enriching the ecosystem and ultimately realizing the path to Ethereum's scaling. We now have a clear development roadmap, much of which has already been implemented or is actively being developed. In the future, I look forward to more exciting achievements, and I am very excited about the current state of the entire Ethereum ecosystem. Thank you.