Overview
Sharding is a scaling technique that splits a blockchain network into smaller, more manageable partitions called shards, allowing transactions to be processed in parallel rather than sequentially. Each shard maintains its own subset of state and transactions, with cross-shard communication handled by the protocol. Sharding was a long-planned approach for Ethereum's scalability roadmap.
How It Works
The network divides accounts and smart contracts across shards, each with its own consensus and validator subset. Shards process transactions in parallel, which multiplies aggregate throughput, then coordinate through cross-shard protocols that ensure global consistency. The main chain or beacon layer coordinates shards and finalizes their state. In practice, sharding adds significant protocol complexity, including the challenge of secure cross-shard transactions.
Why It Matters
Sharding is one of the foundational scaling strategies because it can scale compute and storage, not just throughput. It influenced later designs: Ethereum shifted from sharded execution toward rollup-centric scaling, while other ecosystems, such as NEAR, shipped sharded mainnets. Understanding sharding helps explain why L2 rollups became the preferred scaling path.
Related Concepts
Sharding is an alternative and complement to Layer 2 scaling and fits within Modular Blockchain designs that split data availability and execution. It shares the parallel-processing goal with execution sharding proposals in various Layer 1 roadmaps.