Ethereum activated Dencun to open cheaper data space for rollups

Ethereum activated its Dencun network upgrade on March 13, 2024, at the start of epoch 269568, scheduled for 13:55:35 UTC. Blockscout records Ethereum block 19,426,587 at that timestamp, while contemporaneous reporting said the fork finalized at approximately 14:10 UTC.

The activation mattered because it changed how layer-two rollups could publish transaction data to Ethereum. Dencun’s central feature, EIP-4844, introduced temporary data objects known as blobs and gave them a fee market separate from ordinary execution gas. Rollups could therefore begin replacing comparatively expensive transaction calldata with purpose-built data space.

That was a material protocol milestone, but not a guarantee that every rollup user immediately received lower fees. Each layer-two system still had to update its software, begin submitting blobs and decide how resulting savings would be reflected in user charges.

What Dencun activated

Dencun combined Cancun, the execution-layer changes, with Deneb, the consensus-layer changes. The Ethereum Foundation’s mainnet announcement listed nine Ethereum Improvement Proposals in the package. EIP-4844 received most of the attention because it created blob-carrying transactions and the supporting consensus rules, cryptographic commitments and pricing mechanism.

Under the event-day specification, each blob contained 4,096 field elements of 32 bytes, or 131,072 bytes. The initial parameters targeted three blobs per block and permitted a maximum of six, corresponding to approximately 0.375 megabytes of target data capacity and 0.75 megabytes at the limit. These were protocol limits, not measurements of actual usage on March 13.

Blob data was designed to remain available long enough for rollups and other participants to reconstruct and verify state, without requiring Ethereum nodes to retain it permanently. EIP-4844 required blob sidecars to remain retrievable for at least 4,096 epochs—approximately 18 days under Ethereum’s normal epoch timing. Commitments to the data remained available after the blobs themselves could be pruned.

Why a separate fee market mattered

Before Dencun, rollups generally published compressed transaction data through Ethereum calldata. That competed for conventional blockspace used by other transactions. EIP-4844 gave blob data its own gas accounting and self-adjusting base fee, allowing blob prices to respond to demand independently of ordinary execution gas.

The design addressed a specific constraint in Ethereum’s scaling strategy. Rollups execute transactions outside the base layer, then publish information to Ethereum so their results can be checked. Data publication represented a substantial portion of their costs. A cheaper, specialized channel could reduce that expense while preserving Ethereum as the settlement and data-availability layer.

Dencun was described as proto-danksharding because it introduced the transaction format and much of the supporting machinery anticipated for later scaling work. It did not implement full data sharding or data-availability sampling. Every consensus node still downloaded the blobs, and the conservative initial limits were intended to constrain network load.

The broader upgrade

The eight other proposals addressed developer and validator concerns. They included transient storage opcodes, access to the beacon-block root from the Ethereum Virtual Machine, a memory-copy instruction, revised SELFDESTRUCT behavior, changes to voluntary exits and attestation inclusion, a cap on validator activation churn, and an opcode exposing the blob base fee.

Those changes made Dencun more than a fee-reduction release. It was a coordinated modification of Ethereum’s execution and consensus systems, requiring compatible software across both client layers. The successful activation demonstrated that independently maintained clients could follow the new rules at the agreed fork boundary.

The defensible March 13 conclusion was narrower than the strongest scaling claims: Ethereum had opened a new temporary-data market for rollups and completed the protocol coordination needed to support it. Actual fee reductions, blob demand, rollup adoption and longer-term effects on Ethereum’s economics required measurements after activation and were not established by the fork alone.