Vitalik Buterin Discusses Layered Understanding of Blockchain Scalability

Vitalik Buterin has shared his insights on the layered understanding of blockchain scalability. According to Odaily, he categorizes the difficulty of scaling blockchain components from easiest to hardest as computation, data, and state. Buterin explains that computation is the easiest to scale, achievable through parallelization, using 'hints' provided by block builders, or replacing extensive computation with proofs like zero-knowledge proofs. Data scaling is moderately difficult, especially when systems require data availability guarantees. However, it can be optimized through data splitting, erasure coding methods like PeerDAS, and supporting 'graceful degradation,' allowing nodes with lower data capacity to still generate blocks of corresponding size. In contrast, state is the most challenging aspect to scale. Buterin notes that even to verify a single transaction, nodes require the complete state. Even if the state is abstracted as a tree with only the root node saved, updating this root still depends on the complete state. Although there are methods to split the state, they often necessitate significant architectural changes and are not universally applicable. Based on these observations, Buterin concludes that if data can replace state without introducing new centralization assumptions, it should be prioritized. Similarly, if computation can replace data without new centralization assumptions, it should be seriously considered.