Ethereum co-founder Vitalik Buterin has set out a clear “scalability hierarchy” for blockchains: computation, then data, then state. He explains that each layer has different limits, so engineers should not treat them as equally easy to scale.
According to Buterin, computation is the easiest part to grow. Networks can increase throughput using parallel processing, block-builder “hints,” and zero-knowledge proofs that replace heavy raw computation.
Data sits in the middle of the ladder. When a chain needs strong data availability, it cannot ignore that cost, but it can spread data using sharding, erasure coding, and systems like PeerDAS.
Why Buterin Says State Creates the Toughest Bottleneck
State, which tracks all balances, contracts, and storage, is the hardest element to scale. To fully verify a single transaction, a node usually needs access to the correct, current state.
Buterin notes that even if developers compress state into a tree and keep only the root, updating that root still needs the full underlying data. Some designs can split the state across parts of the network, but they demand big architectural changes and often do not work for every chain.
Because state growth pushes up hardware requirements, it can quietly increase centralization risk. Buterin, therefore, argues that any design that replaces state with data or data with computation without introducing new trusted parties deserves serious attention.
Links to Ethereum’s Current Scaling Path
This hierarchy matches Ethereum’s broader roadmap, which puts rollups and proofs at the center of scaling. Layer 2 networks take on most computation, while Ethereum’s base layer focuses on providing a secure data space and protecting long-term state.
Upgrades like EIP-4844 “blobs,” data-availability sampling, and planned PeerDAS aim to expand data capacity without letting state grow out of control. At the same time, the rise of zkEVMs shows how extra computation and proofs can replace repeated execution on every node.
Projects such as Monad and Ethereum’s upcoming Fusaka changes are already using this logic, optimizing computation and data handling first. By spelling out this ladder, Buterin gives builders a simple rule of thumb: scale with more computation and smarter data, while treating state growth as the last resort
READ MORE: SoFi Stock Price Forms Scary Patterns Ahead of Earnings: Will it Crash?
