Sharding is a method of splitting blockchains (or other types of databases) into smaller, sharded blockchains that process specific segments of data. This setting removes the load on a single chain that handles all transactions and interactions on the network. Each split chain of blocks is known as a shard and has its own special ledger.
These shards can then process their transactions, but the beacon or main chain manages the interaction between the shards. This makes sharding a layer 1 network scalability upgrade as it is a modification of the main blockchain network. Implementing partitions provides much more scalability than continuing to use a single network. Ethereum developers are introducing sharding as they upgrade Ethereum to a scalable Proof-of-Stake network.
In the Ethereum example, the Beacon chain will coordinate 64 separate shards. The original Proof of Work Ethereum chain will merge with the Proof of Stake Beacon chain, with sharding coming later.
One of the benefits of network sharing is the ease and affordability of running a node. Since the network data is distributed across multiple shards, there is no longer a need for validator nodes to store the full history of the entire blockchain. Instead, the validator only needs to store data integrity proofs.
Partitionless networks often resort to using rollups to improve scalability. They accept off-chain transactions and combine them for verification on the main chain. So while sharding may seem like a rival to rollups, it makes them more efficient. Divided network bulletins can report their status more efficiently, which increases their speed.
The main problem with sharding is the possibility of attackers taking over the shard. This fragment can then have a negative impact on other parts of the network. Without proper care and rules, the task of maintaining a fragment is more manageable than maintaining the entire undivided network.
 into smaller, sharded blockchains that process specific segments of data.){kind=link}