Danksharding is Ethereum's planned full sharding design that scales data availability for Layer 2 rollups through data availability sampling.
Sharding involves breaking down a blockchain database into smaller parts for increased efficiency. Each shard has its own blocks and block proposers, but in danksharding, there is a single block proposer per slot and one unified fee market for blob space. This simplifies the architecture while keeping security, decentralization, and scalability as priorities.
The primary focus of danksharding is data availability for rollups, a Layer 2 scaling approach on Ethereum. Rather than requiring every node to download all blob data, the design relies on data availability sampling (DAS): nodes check small random pieces of the data and verify probabilistically that the full dataset is available.
The Road to Full Danksharding
Full danksharding has not shipped yet.
The path toward it has unfolded in stages:
The first step, proto-danksharding (EIP-4844), went live with the Dencun upgrade in March 2024, introducing blob-carrying transactions that gave rollups a dedicated, cheaper data lane instead of competing for regular transaction calldata.
The next step came with the Fusaka upgrade on December 3, 2025, which introduced PeerDAS, a peer-to-peer data availability sampling technique that lets validators verify blob data by sampling small pieces from peers rather than downloading everything. Follow-up Blob Parameter Only (BPO) forks then raised blob capacity from a target of 6 and maximum of 9 blobs per block to a target of 14 and maximum of 21 by January 2026.
These increases have already reduced data costs for rollups, which in turn can lower gas fees for users of Layer 2 networks.
Full danksharding remains the end state of this roadmap: it would raise blob capacity much further (with 64 blobs per block commonly cited as the eventual target) and complete the transition to sampling-based verification, but it has no committed mainnet date as of 2026.
Why Danksharding Matters
Danksharding addresses the scalability side of the blockchain trilemma by allowing Ethereum to serve as a high-throughput data availability layer for rollups, while the rollups handle transaction execution. If fully realized, the design would let the network process a far higher combined transaction volume without raising the hardware requirements for validators, since nodes verify data by sampling rather than downloading. This keeps participation accessible while the network grows, which is the balance the trilemma asks for.