Deconstructing the Ethereum Fusaka Upgrade: The Scalability Evolution Behind 12 EIPs

On June 20, at the 214th Ethereum Execution Layer Core Developers Meeting (ACDE), the core developers agreed to keep the final scope of the Fusaka upgrade largely unchanged, adding only one additional EIP (EIP 7939), which covers 12 EIPs. This also marks the formal transition of Fusaka from the "planning" stage to the "substantive implementation" stage.

As the largest hard fork bundled upgrade since The Merge, the market generally expects that if Fusaka can launch as planned by the end of 2025, it will bring another order of magnitude increase to L2 data space, and L2 transaction fees may further decrease in the next 1-2 years, thereby consolidating Ethereum's position in front of its competitors.

The Continuous Expansion Logic of the Ethereum Roadmap

As we all know, the scalability issue of Ethereum has long been a core bottleneck that kept the mainnet's on-chain costs high and made it difficult for DApps to become popular.

According to the data publicly shared by Vitalik in April this year, the current throughput of Ethereum L1 is 15 transactions per second, and the Gas limit has recently been increased to 36 million, which is about a 6-fold increase over the past 10 years.

At the same time, a more significant transformation is occurring in Ethereum L2, with the current L2 throughput reaching approximately 250 TPS. There have been significant advancements in scalability, and this capability is not just reflected in the data; many users have also clearly felt the reduction in costs and acceleration of on-chain operations.

In the past year, whether it is Arbitrum, Optimism, or Base, L2 transfer fees have generally dropped to the range of 0.01 USD or even lower, achieving a decrease of one or even multiple orders of magnitude compared to before, the daily Gas costs on the Ethereum mainnet have also become significantly more friendly (of course, the impact of market conditions and on-chain activity cannot be ruled out).

This transformation is not accidental, but rather the result of Ethereum strictly following its blueprint and continuously iterating on its roadmap. We can briefly review the key upgrades of the Ethereum network in recent years:

• In 2022, Ethereum successfully transitioned to a PoS mechanism through The Merge, significantly reducing energy consumption and freeing up execution layer bandwidth for subsequent upgrades;
• The Dencun upgrade was successfully activated in 2024, introducing the Blob data mechanism, providing low-cost, temporary storage space for L2, drastically reducing Rollup costs and opening up channels for scalability;
• The recent Pectra upgrade successfully went live on May 7, significantly optimizing the validator operation process, and enhancing the flexibility of participation in the PoS system;

The next step in the Fusaka upgrade is a key step in continuing the above process.

According to the latest statement from Tomasz Kajetan Stańczak, co-executive director of the Ethereum Foundation, Fusaka is set to launch its mainnet in the third or fourth quarter of 2025 (exact timing to be finalized), and plans to implement several core EIPs, including PeerDAS data availability sampling, further propelling Ethereum from performance bottlenecks to mainstream applicability.

It can be said that from The Merge → Dencun → Pectra → Fusaka, Ethereum is steadily advancing towards its long-term blueprint, which is to create a global network that combines security, scalability, decentralization, and sustainability.

Fusaka Upgrade Overview

From the 12 core EIPs included in this upgrade, it basically covers multiple technical dimensions such as data availability, lightweight nodes, EVM optimization, and the collaborative mechanism between the execution layer and the data layer.

Among them, the most notable proposal of this Fusaka upgrade is EIP-7594 (PeerDAS), which introduces a mechanism called "Data Availability Sampling (DAS)" that allows validators in the network to complete verification by downloading only a portion of the Blob data, without needing to store all the data in full.

This greatly reduces the network burden, improves verification efficiency, and paves the way for the large-scale transaction processing capabilities of L2. The concept of "Blob" here can be traced back to EIP-4844, which was introduced in the Dencun upgrade in 2024.

As the most important milestone for Ethereum in 2024, the Dencun upgrade's EIP-4844 enables transactions carrying Blobs for the first time, allowing L2s to choose not to use the traditional calldata storage mechanism, significantly improving the Gas fees required for transactions and transfers on L2.

So what is a carry Blob transaction? In short, it embeds a large amount of transaction data into a Blob, thereby significantly reducing the storage and processing burden on the Ethereum mainnet, not counting towards the Ethereum mainnet state, directly addressing the L1 cost issues related to data availability, ensuring that L2 platforms can offer cheaper and faster transactions without compromising the security and decentralization of Ethereum.

The expansion of Blob here is also based on Pectra — the Pectra upgrade in May increased the Blob capacity from 3 to 6. Notably, Vitalik has publicly stated that ideally, Fusaka will expand the Blob capacity to 72 per block (growing to 12~24 in stages). In the future, if DAS is fully realized, the theoretical maximum capacity could reach 512 Blobs per block.

Once implemented, the processing capacity (TPS) of L2 is expected to soar to tens of thousands, which will greatly enhance the usability and cost structure of high-frequency interaction scenarios such as on-chain DApps, DeFi, social networks, and games. This is also one of the core directions in Vitalik's previously proposed "L2 Security and Finality Roadmap."

At the same time, Fusaka also plans to achieve lightweight state and node structure by introducing Verkle trees, which can significantly compress the size of state proofs, making light clients and stateless validation possible, and also help promote the decentralization of Ethereum and its adoption on mobile devices.

In addition, Fusaka also focuses on the flexibility and performance bottlenecks of the virtual machine layer (EVM), including the following proposals:

• EVM and contract optimization rely on EIP-7939 (CLZ opcode): efficiently implementing bitwise operations to accelerate cryptographic computations;
• EIP‑7951 (secp256r1 alternative support): Enhancing compatibility with Web2 and enterprise architectures;
• EIP‑7907: Expanding the contract size limit to support the deployment of more complex logic, enhancing developer flexibility;

In order to ensure that the expansion does not affect network stability, Fusaka has also introduced EIP-7934 to set block size limits, ensuring that blocks do not become overly heavy due to Blob expansion, and adjusts Blob usage fees through EIP-7892 / EIP-7918 to prevent resource abuse and dynamically match supply and demand fluctuations.

The Watershed for Ethereum Scalability and Experience?

Overall, we will find that Fusaka is not just a technological upgrade, but also has the potential to establish a bridge "from scalability to usability" on multiple key levels.

For Rollup developers, this means lower data writing costs and a more flexible interaction space; for wallet and infrastructure providers, it means supporting more complex interactions and heavier node environments; for end users, it translates to lower experience costs and faster response times for on-chain operations; for enterprises and compliant users, the simplification of EVM scaling and state proofs will also make on-chain interactions easier to integrate with regulatory systems and large-scale deployments.

However, it is still important to remain cautiously optimistic. As of the time of writing, Fusaka is still undergoing testing on multiple Devnets, and the final launch time may still change. In an optimistic scenario, Fusaka is expected to complete its mainnet deployment by the end of 2025, which could become another significant milestone in Ethereum's history following The Merge.

Overall, Fusaka is not only limited to enhancing on-chain scalability but also represents a key step for Ethereum in transitioning to mainstream commercial applications and ordinary users. It is expected to provide a technical foundation for the next phase of the Rollup ecosystem, enterprise-level Dapps, and on-chain user experience.

The true turning point for Ethereum towards large-scale mainstream applications may be approaching.