Ethereum is preparing to take a decisive step in its technical overhaul. As the Fusaka rollout approaches, the network is entering the final testing phase of a key update. Behind this discreet change, however, lies a strategic turning point: laying the foundations for parallel execution, expected in 2026. More than a simple performance gain, Ethereum is initiating a structural transformation, designed to sustainably support its rise in the face of growing scalability challenges.
In brief
- The Fusaka update is entering the final testing phase before its deployment on the mainnet scheduled for December 3, 2025.
- It introduces a gas limit per transaction set at 16.78 million units, to prevent a single transaction from monopolizing an entire block.
- The per-transaction cap allows for better resource allocation, reduces DoS risks, and improves block predictability.
- This update prepares for the arrival of parallel transaction execution, planned for 2026 with the Glamsterdam update (EIP-7928).
A new limit for more stability
The Fusaka update introduces a major new feature: the establishment of a gas cap per transaction set at 16.78 million units, in accordance with the EIP-7825 proposal.
Until now, a single transaction could consume all of the available gas in a block, around 45 million units, putting the network at risk of congestion or denial of service (DoS) attacks. The gas limit per transaction aims to improve block efficiency, reduce DoS risks, and lay the foundation for parallel execution.
This limitation is already active on the Holesky and Sepolia testnets, as part of the preparatory phases for deployment on the main network.
By establishing this limit, Ethereum seeks to improve the resilience of the protocol while streamlining transaction processing. This technical change is also accompanied by an increase in the overall capacity of the blocks, which goes from 45 to 60 million units of gas.
Concretely, this evolution allow :
- To prevent a single transaction from monopolizing the resources of an entire block;
- To encourage the simultaneous inclusion of several lower-intensity transactions;
- To reduce the risks of saturation or targeted attacks on gas consumption;
- To optimize the predictability of block construction for validators and developers.
This overhaul of the internal functioning of the network marks a key step in the professionalization of the Ethereum infrastructure, with a view to accommodating future developments such as parallel execution.
From testnets to mainnet, towards large-scale parallel execution
Beyond this new block architecture, Fusaka introduces a central innovation called PeerDAS, for “ Peer Data Availability Sampling ». This system allows Ethereum nodes to store only a random portion of data “blob” from layers 2, rather than the entire dataset.
The goal is to reduce hardware requirements while maintaining network security. This advance aims to strengthen the adoption of rollup type solutions, which rely precisely on this availability of data. Indeed, this PeerDAS innovation “reduces hardware requirements while maintaining security, and enables more cost-effective, high-throughput scalability for layer 2 networks”.
PeerDAS implementation paves the way for the next step in the Ethereum roadmap: the update “Glamsterdam”expected in 2026, which will introduce EIP-7928, the first concrete iteration of parallel transaction execution. This development will make it possible to process several operations simultaneously, instead of executing them sequentially as is still the case today.
From this perspective, Fusaka acts as an essential prerequisite: without gas limitation per transaction or data-optimized infrastructure, parallel execution could not be deployed securely. For Gabriel Trintinalia, engineer at Consensys,“These testnet updates are essential to build confidence before the mainnet fork”.
Ultimately, this series of improvements, scheduled for December 3, could profoundly transform the internal functioning of Ethereum. The network is moving towards a more modular architecture, more scalable and better adapted to future uses, particularly those of second-layer solutions. They make it possible to validate performance, identify borderline cases and refine technical parameters before activation.
Maximize your Tremplin.io experience with our 'Read to Earn' program! For every article you read, earn points and access exclusive rewards. Sign up now and start earning benefits.
