Sharding
Sharding is a scaling approach that splits a blockchain's data or workload into separate segments, called shards, so that no single participant has to process everything.
The basic idea
In a traditional blockchain, every full node stores and checks every transaction. That keeps things simple and secure, but the whole network can only go as fast as one node. Sharding borrows an idea from databases: divide the work into pieces and let different groups of nodes handle different pieces in parallel.
If a network has, say, 10 shards, each one processes its own share of transactions. In principle, total capacity grows with the number of shards, while each node's workload stays manageable.
Why it is hard
Splitting the network also splits its security. If each shard is checked by only a fraction of validators, an attacker might take over one shard far more cheaply than the whole chain. Designs counter this by assigning validators to shards at random and rotating them frequently.
Transactions that touch several shards, for example a payment from an account on shard 1 to one on shard 7, need messaging between shards. That adds complexity and delays, and it makes building applications harder.
How Ethereum's plan changed
Ethereum originally planned dozens of shard chains that would each execute transactions. It later shifted to a rollup-centric roadmap, where layer 2 rollups do the execution and Ethereum shards only data. The first step, proto-danksharding (EIP-4844), went live in March 2024 and introduced blobs, temporary data space that rollups use to post their batches cheaply.
Full danksharding aims to let nodes check data availability by sampling small random pieces instead of downloading everything. Other networks, such as NEAR, use execution sharding directly.
Why it matters
Sharding is one of the main answers to the blockchain trilemma: adding capacity without forcing every node onto expensive hardware. Whether it succeeds depends on how well a design keeps each shard as secure as the whole.
Ask Coach about it
Coach is the AI on AtenaCrypto. It explains crypto with live market data, in plain words.
How does Ethereum's data sharding help rollups become cheaper?Ask Coach →
Frequently asked questions
Is sharding the same as a layer 2?
No. Sharding splits the base layer itself, while a layer 2 is a separate network built on top of it. Ethereum now combines both: rollups for execution and data sharding for capacity.
Does Bitcoin use sharding?
No. Bitcoin keeps a single chain that every full node verifies, and scales through layers such as the Lightning Network instead.
What is a shard takeover?
An attack in which a malicious group gains control of one shard's validators. Random assignment and frequent rotation make it hard to achieve.
Related terms
Learn it step by step
AC Learning explains these ideas in interactive lessons — the first eight sections are free.
Open AC Learning → Create a free accountAll glossary terms · Educational reference only — not investment, legal, tax or financial advice.