Ethereum's Fusaka upgrade activated on mainnet in July 2026, delivering the network's most consequential scalability release since Pectra. Headline changes include PeerDAS, which increases the number of data blobs each block can carry, and further refinements to the validator effective-balance mechanism introduced under EIP-7251.
Layer-2 networks that batch transactions back to Ethereum, including Base, Arbitrum and Optimism, saw per-transaction fees fall within hours of activation as blob supply expanded. Client teams reported a clean rollout with no reorgs during the first epochs.
Why it matters
Cheaper L2 data is the single biggest lever Ethereum has for keeping stablecoin transfers, on-chain trading and tokenised real-world asset settlements economically viable at institutional scale. With Fusaka live, the network's roadmap turns toward Verkle trees and further stateless-client work.
Trade-offs inside the upgrade
Raising data availability throughput lowers rollup costs, but it also raises the resource floor for running a node. Every scaling decision spends some decentralisation, and the honest debate is about how much.
Client teams have been transparent about the tuning involved. The measurable outcome is whether independent node counts hold up over the next two quarters.
