Danksharding is Ethereum’s approach to expanding data availability for rollups through distributed handling and sampling of blob data. Its focus is giving rollups more space to publish the data needed for verification, rather than creating many independent execution chains for ordinary Ethereum transactions.
Why sampling matters
Requiring every node to download all additional data would increase bandwidth and hardware demands. Data availability sampling lets nodes check selected pieces, while redundancy from erasure coding helps make the original data reconstructible. Cryptographic commitments help verify that received pieces correspond to the published data.
Availability and execution correctness remain different questions. A network can establish that data is obtainable without proving that a rollup calculated its resulting state correctly. That second task still needs the rollup’s verification mechanism.
Implemented steps and the broader roadmap
EIP-4844 introduced blob transactions through Dencun in March 2024. Fusaka introduced PeerDAS in December 2025, extending how nodes distribute and check blob data. As of September 2026, the broader full-Danksharding design remains roadmap work; these deployed steps should not be described as its complete implementation.
More data capacity can reduce a rollup’s data-posting costs, but user fees also depend on demand, execution and the rollup’s own pricing. Blob availability is not permanent archival storage, and a scaling roadmap is not a guarantee of a fixed transaction speed or fee.