EIP-4844 after Dencun: the real impact of blobs on Layer 2

Rollups process transactions outside Ethereum’s base layer and need to publish data that allows their activity to be verified. Publication costs affect the operation of Layer 2 networks. EIP-4844 introduced a dedicated resource for that data: blobs, packages retained temporarily by Ethereum.

The upgrade activated with Dencun on March 13, 2024, separating that publication from the conventional use of calldata, which remains in the transaction history. Understanding its effect requires distinguishing the cost of publishing data to Ethereum from the cost of executing operations within each rollup.

What changed compared with calldata

Before Dencun, a Layer 2 published much of its data as calldata. That resource competes with base-layer execution and forms part of execution history, even though a rollup only needs the data to remain available long enough for anyone to verify or reconstruct its state.

Blobs separate that data from execution space and have their own fee market. When contract activity on Ethereum rises, a Layer 2's data cost no longer has to rise in the same proportion. That separation, rather than any single promotional percentage, is EIP-4844's lasting economic change.

How a blob works

A transaction can attach blobs, but the EVM cannot access their contents directly. The block includes KZG commitments that cryptographically bind each blob to the transaction. Nodes keep the data for a limited period and may then discard it, while the commitment remains part of the verifiable history.

The design uses KZG commitments to prove that data belongs to a committed polynomial without publishing the whole blob inside execution. This prepares the architecture to scale data availability without requiring every validator to retain every rollup batch forever.

The real impact on Layer 2

For a Layer 2, publishing data to Ethereum was one of the dominant costs. Blobs introduced a data resource with its own capacity and price. Savings for a particular operation depend on that price, batch utilisation and how costs are allocated across transactions. A fixed reduction cannot be inferred for every rollup.

A low-cost blob does not automatically mean the lowest possible fee for every user. The final price also includes execution inside the rollup, batch frequency and compression, sequencer margin, Layer 2 congestion and subsidy policy. Comparing networks requires looking at total cost rather than the blob base fee alone.

What EIP-4844 did not solve

EIP-4844 did not by itself increase Ethereum execution capacity, remove bridges or unify liquidity across networks. It also did not implement full danksharding. It introduced the first blob data layer and the cryptographic primitives needed to expand that capacity through later upgrades.

Measuring the effect without mixing costs

A useful comparison observes the same operation type and separates user charges from the operator’s cost. If a rollup changes compression, subsidies and its data resource together, attributing the entire difference to EIP-4844 would be incomplete.

Record the period, transactions per batch, published data, execution fees and publication fees. Include lightly filled batches and higher-demand periods. A favourable average can conceal more expensive individual operations or a temporary commercial policy.

Retaining a commitment must also be distinguished from retaining the data needed to reconstruct historical state. Test archival policy through actual retrieval, with assigned responsibilities and sources available after the network’s required retention period. A cryptographic commitment allows retrieved data to be verified; it cannot recreate an arbitrary blob that nobody retains.

What teams building on rollups should review

Teams should measure execution, data publication and sequencer costs separately; decide when to close each batch; monitor the blob market during spikes; and retain their own indexing and historical-availability strategy. Data expiry at the consensus layer does not remove a protocol's responsibility to keep information its product, proofs or users will need later.

Overall, EIP-4844 validates the rollup-centric roadmap for Ethereum: the base layer provides consensus and data availability, while Layer 2 networks concentrate execution. Its value is not a promise of permanently fixed fees, but a purpose-built, scalable resource for the data that supports those systems.