Points of Focus

• Fusaka upgrade enabled PeerDAS, reducing node bandwidth and opening the path to much higher rollup data capacity.
• Blob-limit increases and a 60M gas cap make Ethereum’s base layer more scalable and predictable for L2s.
• The upgrade brings Vitalik Buterin’s long-planned data-first scaling roadmap to life on mainnet.

Ethereum’s Fusaka hard fork, activated on December 3, 2025, marked one of the network’s most important scaling steps since Dencun. 

The upgrade’s centerpiece, EIP-7594 (PeerDAS), brings peer-based data-availability sampling to the mainnet. Instead of requiring every validator to download complete rollup blob data, nodes can now verify the data by checking small, randomly selected samples. This design lowers per-node bandwidth requirements by roughly eightfold and creates room for far more blob throughput without pushing hardware requirements higher.

Fusaka is live on Ethereum mainnet!

– PeerDAS now unlocks 8x data throughput for rollups
– UX improvements via the R1 curve & pre-confirmatons
– Prep for scaling the L1 with gas limit increase & more

Community members will continue to monitor for issues over the next 24 hrs.

— Ethereum (@ethereum) December 3, 2025

PeerDAS and BPO unlock new data limits for rollups

Blobs, introduced during the March 2024 Dencun upgrade, had already lowered Layer 2 networks’ data costs by over 90%. However, their fixed limits, 6 target and 9 maximum per block, placed a clear cap on rollup throughput. With the Fusaka upgrade, that structural ceiling is gone.

PeerDAS works by splitting each blob into many small pieces, which are spread across the network using erasure coding. Validator nodes only need to sample roughly one-eighth of the pieces to verify that the entire blob is available. This method, also used in leading data-availability research, ensures security even in adversarial conditions because validity depends on cryptographic sampling rather than majority voting.

To scale safely, Fusaka introduces EIP-7892, which enables Blob Parameter Only (BPO) forks. These are small, parameter-level changes that adjust blob limits without a full hard fork. The first two increases are already scheduled:

December 9, 2025: Blob target/max → 10/15

January 7, 2026: Blob target/max → 14/21

For rollup teams building on Arbitrum, Optimism, Base, and newer L2s, these changes immediately reduce costs and increase the reliability of posting data to Ethereum. With blob capacity growing and PeerDAS reducing bandwidth pressure, the L2 environment becomes far more scalable.

Missed the Fusaka network upgrade?
13 Ethereum Improvement Proposals (EIPs) are now live on Mainnet.

Here’s Fusaka in 35 seconds. pic.twitter.com/DlUh1ATA55

— Ethereum (@ethereum) December 4, 2025

Fusaka also lifts Ethereum’s default block gas limit to 60 million, giving the base layer more room for settlement and batch-processing activity. Safety updates, including a new execution block-size cap and strengthened transaction gas constraints, ensure that higher limits do not compromise network stability.

Other improvements, such as a minimum blob base fee (EIP-7918) and a new secp256r1 precompile for wallet interoperability (EIP-7951), update both the economics and cryptography of the protocol.

Together, these changes make Fusaka the first upgrade that actively expands Ethereum’s data capacity, making it truly scalable, rather than just making it cheaper.

Vitalik Buterin’s long-term scaling vision lands on-chain

Fusaka also reflects a clear milestone in the scaling philosophy laid out by Ethereum co-founder Vitalik Buterin. In essays such as the “rollup-centric roadmap” and “Endgame”, Buterin had argued that Ethereum should scale by increasing data availability rather than execution. The chain would provide abundant, verifiable data for rollups, while most user activity happens off-chain on L2 networks.

PeerDAS is the mechanism that turns that vision into reality. By letting validators sample small pieces of blob data to verify full availability, Ethereum can now expand blob capacity many times over while keeping validator hardware requirements modest. This achieves the practical form of sharding that Ethereum researchers had targeted for nearly a decade, without replacing the validator model or compromising decentralization.

Fusaka also prepares the chain for future roadmap phases. With data availability sampling live, Ethereum can increase blob limits through upcoming BPO forks, strengthen proposer-builder separation, and eventually explore higher L1 gas ceilings as zero-knowledge proving systems mature.

For builders and investors tracking Ethereum’s modular roadmap, Fusaka brings a more scalable and predictable data layer, marking the moment the decade-old sharding promise moved from research papers to production mainnet. Rollups have gained breathing room, while the base layer has acquired a proven, decentralized tool for indefinite expansion going forward.

ETH traded at around $3,178 on December 5, 2025, reflecting market confidence in the upgrade’s long-term impact.