Vitalik: The extended hierarchical structure of blockchain consists of computation, data, and state.

PANews reported on January 27th that Ethereum co-founder Vitalik Buterin published an article outlining the scalability hierarchy in blockchain, namely computation, data, and state. He stated that computation is relatively the easiest to scale, achievable through parallel processing, requiring block builders to provide "hints," or directly replacing large amounts of computation with proofs. Data resides in the middle layer; if availability guarantees are required, this requirement cannot be bypassed, but it can be split using techniques such as sharding and erasure coding (e.g., a node can still produce blocks of 1/10th the size when it only has 1/10th the data capacity), and graceful degradation can be achieved (e.g., a node can still produce blocks of 1/10th the size when it only has 1/10th the data capacity).

State is the most difficult element to extend. To ensure the verifiability of a single transaction, the complete state must be obtained. If the state is replaced with a tree and only the root hash is retained, updating the root still requires the complete state. Although there are methods for splitting the state, these methods usually involve architectural changes and lack generality. Therefore, he concludes that if a solution can replace the state with data or replace data with computation without introducing new centralized risks, it should be seriously considered in principle.