Data Availability and Blobs

Data availability asks a simple question with huge consequences: can the network actually access the transaction data needed to verify a claimed state transition? For rollups, this matters as much as execution itself, because a correct proof is not enough if no one can reconstruct what happened.

Before blobs, rollups mostly used calldata, which worked but was expensive. EIP-4844 introduced a cheaper lane for temporary batch data. Blobs are not permanent state like contract storage; they are designed for rollup data that needs to be widely available long enough for verification, then can be pruned.

This is the bridge to the longer-term roadmap. Proto-danksharding delivered the first version of low-cost data posting, while future work such as PeerDAS aims to let validators check larger data loads without requiring massive hardware. In Ethereum's scaling model, abundant data availability is what lets rollups do the high-volume execution.