Ethereum’s surprising usage drop suggests the network solved the wrong problem with Fusaka upgrade

Ethereum activated the Fusaka upgrade on Dec. 3, 2025, elevating the network’s knowledge availability capability by Blob Parameter Overrides that incrementally expanded blob targets and maximums.

Two subsequent changes raised the goal from 6 blobs per block to 10, then to 14, with a most ceiling of 21. The purpose was to scale back layer-2 rollup prices by rising throughput for blob knowledge, the compressed transaction bundles that rollups publish to Ethereum for safety and finality.

Three months into knowledge assortment, the outcomes reveal a spot between capability and utilization. A MigaLabs evaluation of over 750,000 slots since Fusaka’s activation reveals that the network isn’t reaching the target blob count of 14.

Median blob usage truly declined after the first parameter adjustment, and blocks containing 16 or extra blobs exhibit elevated miss charges, suggesting reliability degradation at the edges of recent capability.

The report’s conclusion is direct: no additional will increase in the blob parameter till high-blob miss charges normalize and demand materializes for the headroom already created.

What Fusaka modified and when it occurred

Ethereum’s pre-Fusaka baseline, established through EIP-7691, set the goal at 6 blobs per block with a most of 9. The Fusaka upgrade launched two sequential Blob Parameter Override changes.

The first was activated Dec. 9, elevating the goal to 10 and the most to fifteen. The second was activated Jan. 7, 2026, pushing the goal to 14 and the most to 21.

These adjustments did not require laborious forks, and the mechanism permits Ethereum to dial capability by shopper coordination quite than protocol-level upgrades.

The MigaLabs evaluation, which printed reproducible code and methodology, tracked blob usage and network efficiency throughout this transition.

It discovered that the median blob rely per block fell from 6 earlier than the first override to 4 afterward, regardless of the network’s capability increasing. Blocks containing 16 or extra blobs stay extraordinarily uncommon, occurring between 165 and 259 occasions every throughout the commentary window, relying on the particular blob rely.

The network has headroom it is not utilizing.

One parameter discrepancy: the report’s timeline textual content describes the first override as elevating the goal from 6 to 12, however the Ethereum Foundation’s mainnet announcement and shopper documentation describe the adjustment as 6 to 10.

We use the Ethereum Foundation’s parameters as source: 6/9 baseline, 10/15 after the first override, 14/21 after the second. Nevertheless, we deal with the report’s dataset for noticed utilization and miss-rate patterns as the empirical spine.

Miss charges climb at high blob counts

Network reliability measured by missed slots, that are blocks that fail to propagate or attest appropriately, reveals a transparent sample.

At decrease blob counts, the baseline miss charge sits round 0.5%. Once blocks attain 16 or extra blobs, miss charges climb to 0.77% to 1.79%. At 21 blobs, the most capability launched in the second override, the miss charge hits 1.79%, greater than triple the baseline.

The evaluation breaks this down throughout blob counts from 10 to 21, exhibiting a gradual degradation curve that accelerates previous the 14-blob goal.

This degradation issues as a result of it suggests the network’s infrastructure, reminiscent of validator hardware, network bandwidth, and attestation timing, struggles to deal with blocks at the higher finish of capability.

If demand finally rises to fill the 14-blob goal or push towards the 21-blob most, the elevated miss charges might translate into significant finality delays or reorg danger. The report frames this as a stability boundary: the network can technically course of high-blob blocks, however doing so persistently and reliably stays an open query.

Blob economics: why the reserve worth flooring issues

Fusaka did not solely develop capability. It additionally modified blob pricing by EIP-7918, which introduces a reserve worth flooring to stop blob auctions from collapsing to 1 wei.

Before this modification, when execution prices dominated and blob demand stayed low, the blob base price might spiral downward till it successfully disappeared as a worth sign. Layer-2 rollups pay blob charges to publish their transaction knowledge to Ethereum, and people charges are purported to replicate the computational and network prices that blobs impose.

When charges fall to close zero, the financial suggestions loop breaks, and rollups eat capability with out paying in proportion. This ends in the network shedding visibility into precise demand.

EIP-7918’s reserve worth flooring ties blob charges to execution prices, making certain that even when demand is delicate, the worth stays a significant sign.

This prevents the free-rider problem the place low cost blobs encourage wasteful usage and gives clearer knowledge for future capability choices: if blob charges keep elevated regardless of elevated capability, demand is real; in the event that they collapse to the flooring, headroom exists.

Early knowledge from Hildobby’s Dune dashboard, monitoring Ethereum blobs, reveals that blob charges have stabilized after Fusaka quite than persevering with the downward spiral seen in earlier durations.

The common blob rely per block confirms MigaLabs’ discovering that utilization hasn’t surged to fill the new capability. Blocks routinely carry fewer than the 14-blob goal, and the distribution stays closely skewed towards decrease counts.

What the knowledge reveals about effectiveness

Fusaka succeeded in increasing technical capability and proving the Blob Parameter Override mechanism works with out requiring contentious laborious forks.

The reserve worth flooring seems to be functioning as meant, stopping blob charges from changing into economically meaningless. But utilization lags behind capability, and reliability at the edges of recent capability reveals measurable degradation.

The miss charge curve suggests Ethereum’s present infrastructure comfortably handles the pre-Fusaka baseline and the first override’s 10/15 parameters, however begins to pressure previous 16 blobs.

This creates a danger profile: if layer-2 exercise surges and pushes blocks towards the 21-blob most frequently, the network might face elevated miss charges that compromise finality and reorg resistance.

Demand patterns supply one other sign. Median blob usage falling after the first override, regardless of elevated capability, suggests that layer-2 rollups aren’t at the moment constrained by blob availability.

Either their transaction volumes have not grown sufficient to require extra blobs per block, or they’re optimizing compression and batching to suit inside present capability quite than increasing usage.

Blobscan, a devoted blob explorer, reveals particular person rollups posting comparatively constant blob counts over time quite than ramping as much as exploit new headroom.

The pre-Fusaka concern was that restricted blob capability would bottleneck Layer 2 scaling and maintain rollup charges elevated as networks competed for scarce knowledge availability. Fusaka addressed the capability constraint, however the bottleneck seems to have shifted.

Rollups aren’t filling the accessible house, which suggests both demand hasn’t arrived but or different elements, reminiscent of sequencer economics, consumer exercise, and cross-rollup fragmentation, are limiting progress greater than blob availability was.

What comes subsequent

Ethereum’s roadmap consists of PeerDAS, a extra elementary redesign of knowledge availability sampling that may additional develop blob capability whereas bettering decentralization and safety properties.

However, the Fusaka outcomes counsel that uncooked capability is not the binding constraint proper now.

The network has room to develop into the 14/21 parameters earlier than needing one other enlargement, and the reliability curve at high blob counts signifies that infrastructure upgrades might must catch up earlier than capability will increase once more.

The miss charge knowledge gives a transparent boundary situation. If Ethereum pushes capability larger whereas 16+ blob blocks nonetheless present elevated miss charges, it dangers introducing systemic instability that might floor throughout high-demand durations.

The safer path is to let utilization rise towards the present goal, monitor whether or not miss charges enhance as shoppers optimize for larger blob hundreds, and regulate parameters solely as soon as the network demonstrates it may well reliably deal with edge instances.

Fusaka’s effectiveness relies on the metric. It expanded capability efficiently and stabilized blob pricing by the reserve flooring. It did not drive fast utilization will increase or resolve the reliability challenges at most capability.

The upgrade created headroom for future progress, however whether or not that progress materializes stays an open query the knowledge hasn’t answered but.

The publish Ethereum’s surprising usage drop suggests the network solved the wrong problem with Fusaka upgrade appeared first on CryptoSlate.