Decentralized networks continue to evolve and find applications in various fields, including finance, gaming, logistics, and even art. But as the load increases, many blockchains cannot meet the growing needs of developers and users — fees rise, and transaction confirmation times increase. This limits the application of the technology and forces projects to seek alternatives.
To eliminate these barriers, projects are developing scaling solutions. One of them is sharding.
What is sharding?
Sharding (sharding) is the division of large volumes of data or tasks into smaller and independent units. The idea of sharding itself is not new and is used where it is important to increase performance without increasing the demands on each individual participant in the network.
As the volume of information and the number of operations grow, traditional systems hit technical limitations — at some point, it becomes challenging to process all requests. Data separation allows several servers to be used, each processing only its part.
For example, Notion uses this technology to separate data between workspaces, and X (formerly Twitter) uses it to handle billions of messages.
In blockchain, this concept has developed as a way to increase throughput and transaction processing speed.
Why is this technology important for blockchain
Modern blockchains face scalability limitations. In most networks, all transactions undergo sequential processing within a single stream, so as the load increases, fees rise, and confirmation times increase.
For scaling, solutions like rollups are used — Zero-Knowledge Rollups and Optimistic Rollups process transactions off the main network and then record their results in the blockchain.
However, these approaches have architectural limitations and do not eliminate the bottlenecks of the underlying infrastructure. Sharding is considered a first-level solution that does not require the involvement of third-party networks.
The essence is that instead of sequential processing, the blockchain is divided into independent or semi-independent segments, each of which performs its part of the work. This allows validators to be utilized more effectively and distribute the load without abandoning decentralization and security.
How sharding works
There are several ways to implement sharding. They can be conditionally divided based on two criteria: what exactly is shared between nodes and how it is organized. Depending on which aspects of operation are segmented, the following can be distinguished:
transaction sharding — the common stream of transactions is divided into several smaller ones, and then each segment of the network processes its part and keeps a local record. This increases the network's throughput;
state sharding — separates the blockchain data between segments. Each processes only its part of the information. This reduces the load on storage and computation, but due to fragmentation, the risk of attacks on individual shards increases, and interaction becomes more complex.
At the same time, there are also two main approaches to network sharding, differing in implementation complexity and flexibility:
static division — assumes segmenting the state of the blockchain into a predetermined number of shards and nodes in each of them. This architecture is simpler to implement but does not adapt to changes in user activity, limiting efficiency;
adaptive separation — a more flexible approach in which the number or size of shards changes based on the number of active nodes, data volume, and network activity. If the load increases, the system can automatically add new segments to handle additional transactions.
Depending on the tasks and architecture, a blockchain can combine several types of sharding simultaneously. This helps achieve the desired balance between performance, decentralization, and security.
Sharding at a technical level
The most common in the crypto industry is transaction sharding. In its simplest implementation, this approach assumes that the entire stream of operations is divided into segments and then distributed among the shards — each with its own group of validators. A kind of 'mini-blockchain'.
Transactions related to one shard are processed only within it. Validators collect them and then, through interaction with other shards, reach consensus on the block and add it to the network.
If validators from different shards cannot interact, such segments essentially turn into isolated networks. To avoid this, inter-shard (cross-shard) transactions are used. They can be:
synchronous. In this case, the transaction is recorded simultaneously in two shards, and validators from several segments act consistently to ensure data integrity. An example of such an approach is the concept of Merge Blocks;
asynchronous. Each part of the transaction is executed independently: the first shard initiates the operation, and the second completes it when it ensures that the first part has been executed. This method is technically simpler but requires additional security mechanisms.
Typically, to ensure communication between segments, a separate coordination layer is provided in the architecture of a sharded network. In Ethereum, for example, this role is performed by the Beacon chain. This layer can perform various functions:
distribute validators across segments;
add or remove shards when the load changes;
synchronize their work;
check the integrity and correctness of calculations.
The structure with a coordination level allows maintaining the integrity and security of the network even when divided into dozens of independent segments. However, it requires protection for both individual shards and the main chain, as well as data exchange mechanisms.
Which projects use sharding?
Relatively few networks use sharding technology or are in the process of implementing it. Their examples show that the concept is being realized and developed in various variations, and a unified standard has yet to be established.
NEAR Protocol
#NEAR🚀🚀🚀 Protocol implements sharding through the Nightshade model, where the network is divided into independent segments, each of which processes its part of transactions. All blocks are signed by a common group of validators working based on the Proof-of-Stake mechanism, ensuring data integrity.
Thanks to this approach, $NEAR can scale linearly — one shard processes about 1000 transactions per second. For example, in March 2025, NEAR expanded from 6 to 8 shards.
The network also implements the single-block resharding technology. It allows splitting an existing segment into two within a single block. This increases configuration flexibility and brings the network closer to dynamic sharding — where segments can be automatically added as the load grows.
Furthermore, NEAR validators are not required to store complete information about other shards. Instead, they use special proofs, which allows the network to scale not only at the transaction processing level but also at the data availability level.
MultiversX
MultiversX has been operating since July 2020, using adaptive state sharding. The network is divided into three segments and a common chain (Metachain), where block finalization occurs. Validators are regularly redistributed among shards, increasing decentralization and security.
MultiversX processes about 240-250 thousand transactions per day, and the number of active addresses exceeds 110 thousand. The claimed maximum performance is up to 15,000 transactions per second, but the actual load is significantly lower.
Sui
$SUI also uses the principle of process separation to enhance scalability; however, their approach differs from the ones described above. Thus, the implemented Pilotfish model breaks down transaction processing between three types of nodes:
Primary — responsible for ordering transactions;
Sequencing Workers (SW) — distribute transactions for execution;
Execution Workers (EW) — store state and execute transactions.
This allows for linear scaling of the network's throughput. In tests, the Pilotfish prototype showed performance growth in direct proportion to the number of nodes involved. Thus, when using eight servers, performance increased eightfold, while the latency per transaction did not exceed 20 ms.
Aptos
In early 2025, the team #Aptos introduced the Shardines solution, in which the transaction block is split into parts and processed in parallel using Executor Shards. In tests, the technology demonstrated scalability up to 1 million transactions per second with low latency, but the data availability level remains a bottleneck.
Currently, Shardines is not used in the main network, but Aptos is working on a Jellyfish Merkle Tree for nodes that will solve the data storage problem. In the future, it is planned to shard all key components of the network.
It is worth noting separately #Ethereum , which is developing its own segmentation model called danksharding. At the time of writing, only the first stage has been implemented — proto-danksharding (EIP-4844), introduced in March 2024 along with the Dencun update. It aims to prepare the infrastructure for more massive future upgrades.
Advantages and disadvantages of the technology
As a method of increasing blockchain performance, sharding has several key advantages over alternative approaches:
scalability and performance. Segmentation allows more transactions to be executed through the parallel operation of the network, increasing overall throughput;
reduced load on nodes. Since each node works only with a part of the data, sharding reduces the computational resource requirements. This lowers the threshold for running nodes and potentially contributes to decentralization;
isolation of overloads. Dividing the blockchain into segments allows for localized load — even if one shard is overloaded, others can operate normally. This approach makes the network more resilient to spikes in activity;
potential for lower fees. With increased throughput and more efficient data storage, sharding contributes to reducing transaction costs.
Nevertheless, sharding is not an 'industry standard' and is not used in all networks. The reasons for limited adoption include:
the complexity of implementation. Full-fledged sharding is one of the most complex tasks in blockchain development. Its implementation requires a team and the development of new solutions for synchronization, protection, and decentralization of individual segments of the network;
security threats. Dividing the blockchain into parts can potentially simplify attacks since, instead of compromising the entire blockchain, an attacker can focus on one segment with fewer validators. This increases the risk of capturing an individual shard, although it is unlikely to lead to an attack on the entire network;
the complexity of inter-shard coordination. Executing or storing transactions across different segments requires synchronization. This can increase the time taken to process operations, complicate the architecture, and raise the likelihood of technical failures;
the absence of a standard. Many blockchain sharding models remain experimental. Different networks implement it in their own way, so there are simply no universal or standardized solutions. The technology requires long testing and audits.
Thus, segmentation enables the creation of scalable and decentralized blockchains, but it is associated with technical issues and complex implementation.
Prospects for the development of sharding in the crypto industry
As a way of scaling blockchains, sharding likely has the greatest potential compared to other solutions. Therefore, despite the technical complexity, interest in sharding is growing, and in the future, it may become the standard architecture for new networks.
At the same time, there is no universal model yet. Each network experiments with implementation: from the number of shards to the methods of synchronization and load distribution. Therefore, the viability of existing projects will show whether sharding can gain wider adoption.
In the future, this technology will likely be applied by universal blockchains focused on decentralization and security.
Conclusions
Sharding is one of the ways to enhance the efficiency of networks, which is already being applied in several projects. The mechanism provides a noticeable increase in scalability while maintaining the decentralization and security of the blockchain. With proper implementation, segmentation helps cope with overload and create a more resilient infrastructure.
At the same time, the technology is still far from mass adoption. Most projects face technical challenges, and implementation requires time, testing, and the development of new, often unique solutions. The future of sharding depends on how successfully current limitations can be addressed.
