Ethereum Usage Plummets: Did the Fusaka Upgrade Miss the Mark?

Ethereum’s recent Fusaka upgrade, activated on December 3, 2025, aimed to significantly boost the network’s data availability capacity through Blob Parameter Overrides. The intention was clear: reduce layer-2 rollup costs by increasing throughput for blob data – the compressed transaction bundles crucial for Ethereum’s security and scalability. However, three months post-activation, data reveals a surprising trend: utilization hasn’t kept pace with the expanded capacity. This raises a critical question – did the Fusaka upgrade truly hit its target, or has it created headroom without a corresponding surge in demand? This article dives deep into the data, analyzing the upgrade’s impact, identifying potential bottlenecks, and exploring what the future holds for Ethereum’s data availability.

What Changed with the Fusaka Upgrade?

Prior to Fusaka, Ethereum’s baseline, established by EIP-7691, allowed for 6 blobs per block, with a maximum of 9. The Fusaka upgrade introduced two sequential Blob Parameter Override adjustments, implemented without requiring disruptive hard forks. This innovative mechanism allows for dynamic capacity adjustments through client coordination, offering a flexible approach to network optimization.

• December 9, 2025: First override – Target increased to 10 blobs per block, maximum to 15.
• January 7, 2026: Second override – Target increased to 14 blobs per block, maximum to 21.

The MigaLabs analysis, utilizing reproducible code and methodology, meticulously tracked blob usage and network performance throughout this transition. Their findings paint a picture that diverges from initial expectations.

Utilization Lags Behind Capacity: The MigaLabs Report

Contrary to expectations, the median blob count per block decreased from 6 before the first override to just 4 afterward, despite the substantial capacity expansion. Blocks containing 16 or more blobs remain exceedingly rare, occurring only between 165 and 259 times across the observation window. This indicates a significant gap between the network’s potential and its actual usage.

The report highlights a concerning trend: the network possesses considerable headroom that remains largely untapped. This raises questions about the factors limiting layer-2 rollup adoption of the increased capacity.

Miss Rates Climb at Higher Blob Counts

Network reliability, measured by missed slots (blocks failing to propagate or attest correctly), reveals a clear correlation between blob count and performance. While baseline miss rates hover around 0.5% at lower blob counts, they escalate significantly as blocks approach maximum capacity.

• 16+ Blobs: Miss rates climb to 0.77% - 1.79%.
• 21 Blobs (Maximum Capacity): Miss rate reaches 1.79%, more than triple the baseline.

This degradation suggests that the existing infrastructure – including validator hardware, network bandwidth, and attestation timing – struggles to reliably handle blocks at the upper end of the capacity spectrum. If layer-2 activity were to surge and consistently push blocks towards the 21-blob maximum, the elevated miss rates could translate into finality delays and increased reorg risk. This represents a critical stability boundary for the network.

The Role of EIP-7918 and the Reserve Price Floor

Fusaka wasn’t solely about expanding capacity; it also introduced EIP-7918, establishing a reserve price floor for blobs. This crucial change aimed to prevent blob auctions from collapsing to 1 wei, a scenario that previously undermined the economic signaling mechanism.

Before EIP-7918, low demand and dominant execution costs often drove blob fees to near zero, effectively eliminating a vital price signal. Layer-2 rollups pay these fees to post transaction data to Ethereum, and these fees should accurately reflect the computational and network costs associated with blob processing. The reserve price floor ensures that even during periods of low demand, blob fees remain meaningful, preventing free-riding and providing clearer data for future capacity decisions.

Early data from Hildobby’s Dune dashboard confirms that blob fees have stabilized post-Fusaka, avoiding the downward spiral observed previously. However, the average blob count per block remains consistently below the 14-blob target, reinforcing the findings of MigaLabs.

What Does the Data Reveal About Fusaka’s Effectiveness?

The Fusaka upgrade demonstrably succeeded in expanding Ethereum’s technical capacity and validating the Blob Parameter Override mechanism as a non-disruptive method for network optimization. The reserve price floor appears to be functioning as intended, preventing economically meaningless blob fees.

However, the lack of corresponding utilization growth and the observed reliability degradation at higher blob counts present a nuanced picture. The miss rate curve suggests that the network comfortably handles the pre-Fusaka baseline and the initial 10/15 parameters, but begins to strain beyond 16 blobs.

This creates a potential risk profile. A surge in layer-2 activity pushing blocks towards the 21-blob maximum could lead to elevated miss rates, compromising finality and increasing the risk of network reorganizations. The current priority should be to allow utilization to rise towards the existing target, monitor miss rates for improvement as clients optimize for higher blob loads, and adjust parameters only when the network demonstrates reliable handling of edge cases.

Looking Ahead: PeerDAS and Future Capacity Expansion

Ethereum’s roadmap includes PeerDAS, a more fundamental redesign of data availability sampling. This ambitious project promises to further expand blob capacity while simultaneously enhancing decentralization and security. However, the Fusaka results suggest that raw capacity isn’t the primary constraint at this moment.

The network has ample room to grow into the existing 14/21 parameters before requiring another expansion. Furthermore, infrastructure upgrades may be necessary to address the reliability challenges observed at higher blob counts before further capacity increases are considered. The miss rate data provides a clear boundary condition for future development.

In conclusion, Fusaka was a successful upgrade in terms of expanding capacity and stabilizing blob pricing. However, it didn’t immediately drive utilization increases or resolve the reliability concerns at maximum capacity. The upgrade has created valuable headroom for future growth, but whether that growth materializes remains an open question, one that ongoing data analysis will continue to illuminate.

Mentioned in this article

Ethereum Dune Analytics Ethereum Foundation