I just saw that Vitalik has detailed the expansion plan for Ethereum over the next five years, which is quite interesting because he’s not only talking about technical aspects but also proposing a step-by-step architectural vision.

According to his statement, expanding Ethereum involves managing three types of resources: execution resources, data resources, and state resources. Each has different plans for the short and long term.

For execution resources, which relate to the computation of the EVM—the virtual machine used to process transactions on Ethereum—short-term plans aim to improve efficiency by approximately 10-30 times through the introduction of Block Access Lists (BAL) and ePBS, which will allow validators to process multiple transactions simultaneously instead of one by one. In the long term, ZK-EVM could boost performance up to 1,000 times.

Another interesting point is multi-dimensional gas pricing. His idea is that different operations should have different costs, especially for creating new states (such as creating a new account), which should have a higher fee because it’s a costly operation and uses permanent resources. He proposes using a reservoir mechanism to make regular transactions cheaper but increase the cost for creating new states.

For data resources, short-term improvements include a 10-20x efficiency boost through enhancements in P2P and multi-dimensional gas. In the long term, Blobs and PeerDAS will help increase efficiency by approximately 500 times.

The most challenging issue is state resources. Currently, Ethereum’s state size is about 100 GB, but expanding it 20 times would result in 2 TB, and eventually 8 TB. The problem isn’t storage space but database performance. As data grows, write speeds slow significantly, and new nodes will take a very long time to download the entire state.

To address this, Vitalik has proposed several new state models, such as temporary storage that expires automatically (like clearing data every month), which can be used for specific data like order books or liquidity pools. There are also models for periodic storage and limited access storage, which can only be accessed through specific methods.

This concept is quite innovative because it offers developers options: they can continue using the existing state model but pay higher fees, or design new applications with new state models to save on transaction costs. Ordinary users benefit from lower fees, while developers need to be more creative in their design.

Overall, Vitalik’s plan seems to be an effort to balance performance improvements with the long-term sustainability of Ethereum.