Ethereum sharding is the network upgrade designed to resolve the scalability problem.

Once implemented, it will enable the Ethereum network to process a much higher number of transactions per second and, at the same time, reduce ETH gas fees.

The long-awaited solution still looms on the horizon with no exact date for implementation. Nonetheless, it is important to investigate the Ethereum sharding technology now so as to make use of it at its fullest upon release.

What is Ethereum Sharding?

Ethereum sharding is a family of scaling upgrades designed to increase the network's capacity by separating how data is stored and verified from how transactions are executed.

The current approach, called Danksharding, focuses on data availability rather than splitting the EVM into independent execution shards. Instead of assigning validators to separate chains, Danksharding introduces blob-carrying transactions — large packets of cheap, temporary data that rollups (Layer 2 networks) can use to post transaction data to Ethereum at a fraction of the previous cost.

This approach is being rolled out in stages. Proto-Danksharding (EIP-4844), the first step, introduced blob transactions to Ethereum mainnet. The Fusaka upgrade in February 2026 built further on this blob framework. Full Danksharding, which adds data availability sampling (DAS) and scales blob capacity significantly, remains in active development as part of Ethereum's multi-phase roadmap.

The Problems of Ethereum Architecture

The Ethereum mainnet relies on a distributed network of nodes that support the network. Each Ethereum node stores information about all transactions that have ever happened within the network since its launch.

Initially, Ethereum was working on the proof of work consensus mechanism. It implies solving complex computational tasks for the new nodes to be created.

In accordance with the development roadmap, the network launched a PoS-based Beacon Chain on December 1st, 2020. Finally, in September 2022, the Merge event took place as the whole network switched to PoS.

Thus, Ethereum has transferred to the proof of stake consensus and overcome the problem of excessive electricity consumption inherent to PoW-based systems. Yet, the system remained unscalable.

At the time of writing, it already features comparatively low fees and fast transaction speed. Yet, it’s still not ready to withstand high network congestion.

With the advance of Ethereum 2.0, the system should implement sharding so as to solve its last problem.

How Does Ethereum Sharding Work?

Ethereum's sharding strategy has evolved significantly. Rather than splitting the network into separate chains where validators process different transactions, the current design focuses on data availability sharding through blobs and Data Availability Sampling (DAS).

Here is how the current phased approach works in practice:

Step 1: Blob-Carrying Transactions (Proto-Danksharding / EIP-4844)

Rollups (like Optimism and Arbitrum) need to post transaction data back to Ethereum's Layer 1 to inherit its security. Before EIP-4844, they posted this data as expensive calldata. Blob transactions introduce a new, cheaper data type. Each blob holds approximately 128 KB of data. Blob data is attached to blocks but is NOT executed by the EVM, keeping it cheap. Critically, blob data is ephemeral: it can be deleted from nodes after roughly 18 days once data availability is confirmed, preventing unbounded storage growth.

Step 2: Data Availability Sampling (DAS)

Instead of every validator downloading every blob in full, DAS allows validators to confirm data availability by sampling small random chunks. If enough validators confirm their samples are available, the full data is considered available to the network. This means even nodes with limited bandwidth can participate in securing data availability.

Step 3: Full Danksharding

Full Danksharding scales blob capacity up to 64 blobs per block (compared to the initial small number introduced in Proto-Danksharding), distributed across the network using DAS. This is the stage that enables massive throughput increases for rollups and the broader Ethereum ecosystem.

The Ethereum Sharding Roadmap: Phase by Phase

Ethereum's sharding is not a single upgrade but a staged rollout. Here is where each phase stands today:

👉 Quick takeaway: EIP-4844 (Proto-Danksharding) is live and already reducing L2 data costs. The Fusaka upgrade expanded the blob framework in February 2026. Full Danksharding — targeting up to 64 blobs per block — remains in active development.

Phase	Upgrade / EIP	Status	Key Change	Blobs Per Block
Phase 0	Proto-Danksharding `EIP-4844`	🟢 Live March 2024	Introduced blob transactions; cheaper L2 data posting 🏆 First live reduction in L2 data costs	3 target / 6 max (initial)
Phase 1	Fusaka Upgrade	🟢 Live February 2026	Expanded blob framework for ETH stakeholders	Incremental increase
Phase 2	Full Danksharding	⚠️ In Development	DAS at scale; horizontal blob sharding across validators 🏆 Target end state for L2 scalability	Up to 64 blobs per block
Phase 3	Cross-Shard Execution	⚠️ Future Research	Cross-shard composability and state sharding	TBD

What This Means for You Right Now

If you use Layer 2 networks (Arbitrum, Optimism, Base): EIP-4844 already reduced your transaction fees by enabling cheaper data posting to L1.
If you are a validator: DAS will change your data verification responsibilities as full Danksharding rolls out.
If you hold ETH: Each phase is designed to increase network utility and throughput without breaking existing applications.

Advantages of Ethereum Sharding

The new approach can help to solve a number of problems inherent to the Ethereum blockchain. At this, Ethereum sharding comes with the following advantages:

Improved scalability for rollups. The primary scalability benefit of Danksharding is dramatically increasing the data throughput available to Layer 2 rollups, not directly increasing L1 TPS. By scaling from a small initial number of blobs per block (Proto-Danksharding) to up to 64 blobs per block (full Danksharding), Ethereum enables rollups to post far more transaction data to L1 cheaply. The practical result is lower fees and higher capacity on the L2 networks where most users transact.
Reduced gas fees. Sharding will also resolve another key pain of Ethereum users, i.e. high gas fees. To verify transactions, users won’t have to compete for the network resources so hard. Therefore, the fees will be much lower.
Reduced storage burden via ephemeral blob data. Under Danksharding, blob data attached to blocks does not need to be stored permanently by every node. Once data availability sampling confirms the data was accessible, blob data can be deleted from local storage. This keeps the long-term storage requirements for running a node manageable even as transaction volume grows significantly.
Higher decentralization. As setting up a node will be much easier, more validators will be able to join the network. The higher their number, the higher the overall decentralization, and, consequently, the higher the security of the whole system.

Disadvantages of Ethereum Sharding

While the advantages of Ethereum sharding are obvious, there are also some concerns to address as well:

Higher complexity. Sharding makes the whole blockchain infrastructure more complex. This, in turn, increases the costs of development and makes vulnerabilities more feasible.
A comparatively new technology. Although shading itself has been around for a while already, it is still a novelty in the blockchain industry. At this, it hasn’t been properly tested yet and, therefore, some unknown vulnerabilities may emerge in time.

How Does Ethereum Sharding Address Security Concerns?

The Danksharding design addresses security through several mechanisms that differ from the older shard-chain model:

Data Availability Sampling (DAS) as the primary security layer. Rather than requiring every validator to download every blob in full, DAS allows validators to confirm data availability by sampling random small chunks. If a malicious actor tried to withhold data, the probability of passing DAS checks without having the full data available is cryptographically negligible. This makes data withholding attacks extremely difficult without requiring every node to store everything.
Random validator committee assignment. Validators are still randomly assigned to verification duties, making it statistically very difficult for a coordinated group to control enough of the sampling process to fake availability.
Ephemeral data with confirmed availability. Because blob data is only deleted after DAS has confirmed availability across the network, there is a cryptographic guarantee that the data existed and was accessible before any deletion occurs. This prevents a scenario where data is posted and immediately hidden.

Note: The 51% attack risk on individual shards cited in older analyses is less relevant under the current DAS-based model, since no single 'shard' processes independent transactions — all execution remains on the unified L1 chain.

Ethereum Sharding Timeline: What Has Happened and What Is Next

Ethereum's sharding rollout is no longer just on the horizon. Significant milestones have already been deployed:

March 2024: Proto-Danksharding (EIP-4844) goes live. Blob-carrying transactions are introduced to Ethereum mainnet, immediately reducing data posting costs for Layer 2 rollups.
February 2026: The Fusaka upgrade is deployed, building on the blob framework established by EIP-4844 and advancing the data availability infrastructure for ETH stakeholders.
2026 Roadmap: The Ethereum Foundation's three-track 2026 protocol roadmap continues to advance blob and data-sharding work alongside post-quantum security considerations.
Full Danksharding: Active development continues. The Ethereum research community is refining data availability sampling architecture and sharded blob mempool designs. No firm deployment date has been announced, but the phased approach means the network is already delivering partial benefits today.

The key shift in framing: sharding is no longer a single future event to wait for. Each phase delivers real benefits. Proto-Danksharding already lowered L2 fees. Full Danksharding will scale this further.

What Does Ethereum Sharding Mean for You?

Ethereum's sharding progress affects different participants in different ways. Here is a practical breakdown:

If you are an L2 user (Arbitrum, Optimism, Base, etc.):

EIP-4844 already reduced the cost of rollup data posting to Ethereum L1. This has translated to lower transaction fees on major L2 networks. As full Danksharding scales blob capacity to 64 blobs per block, L2 fees are expected to decrease further.

If you are an Ethereum validator:

Under full Danksharding, your role in data availability sampling will change. Instead of downloading full blob data, you will be required to sample random chunks of blob data to confirm availability. This reduces per-validator bandwidth requirements compared to downloading everything, enabling more validators to participate.

If you are a dApp developer:

Blob data is not accessible to the EVM, so it cannot be read by smart contracts. Blobs are specifically designed for rollup data posting, not for on-chain computation. Your smart contract logic on L1 is unaffected by sharding. The benefit comes from cheaper and faster L2 environments where your users transact.

If you are an ETH holder:

Each sharding phase is designed to increase Ethereum's utility and network capacity without disrupting existing operations. Greater throughput and lower fees on L2 networks increase overall network demand and usage.

Frequently Asked Questions About Ethereum Sharding

What is the difference between Proto-Danksharding and full Danksharding?

Proto-Danksharding (EIP-4844) introduced blob-carrying transactions with a small initial number of blobs per block. It is already live on Ethereum mainnet. Full Danksharding scales this to up to 64 blobs per block and adds data availability sampling (DAS), allowing validators to confirm blob availability without downloading every blob in full. Full Danksharding is currently in development.

What is a blob in Ethereum?

A blob (Binary Large Object) is a new type of data packet introduced by EIP-4844. Blobs carry roughly 128 KB of data each, are cheaper than regular calldata, and are designed for rollups to post their transaction data to Ethereum L1. Blob data is not accessible to the EVM and is ephemeral: it can be deleted from nodes after approximately 18 days once data availability is confirmed.

How does Danksharding reduce gas fees?

Danksharding reduces fees specifically for Layer 2 rollups, not directly for L1 transactions. By providing a cheaper and larger data lane (blobs) for rollups to post their transaction data to L1, rollups can pass those savings on to their users in the form of lower transaction fees.

What is data availability sampling (DAS)?

DAS is a technique that allows validators to confirm that blob data is available on the network without each validator downloading the full blob. Each validator samples small random chunks of the data. If enough validators confirm their samples, the network considers the full blob available. This enables the network to scale blob capacity without proportionally increasing validator bandwidth requirements.

Is Ethereum sharding live yet?

Partially, yes. Proto-Danksharding (EIP-4844) went live in March 2024, and the Fusaka upgrade in February 2026 continued building out the blob framework. Full Danksharding with complete DAS implementation remains in active development with no confirmed deployment date.

The Bottom Line

Ethereum sharding is no longer a distant promise. Proto-Danksharding (EIP-4844) is already live, blob transactions are already reducing Layer 2 fees, and the Fusaka upgrade (February 2026) has continued to build out the blob data infrastructure.

Full Danksharding, which will scale blob capacity up to 64 blobs per block with data availability sampling, remains in active development. The Ethereum Foundation's 2026 roadmap confirms that blob and data-sharding work continues alongside post-quantum security improvements.

For most Ethereum users, the practical benefit of sharding is already arriving incrementally through cheaper L2 transactions. The full vision of massively scalable, low-cost Ethereum is being built phase by phase rather than delivered in a single upgrade.

Stay current with Ethereum's sharding progress by following the official Ethereum roadmap.

What Is Ethereum Sharding? How Danksharding Works