In June this year, Ethereum core developers gathered offline in Berlin with technical teams from Layer 2 and ZK scenarios for a five-day deep collaboration. This meeting, called Berlinterop, was not only an engineering promotion gathering but also showcased three clear technology roadmaps for Ethereum towards 2025: accelerated execution layer upgrades, restructured inter-chain collaboration mechanisms, and the standardized transformation of zero-knowledge paths.
This is Berlinterop — an offline R&D gathering that continues the tradition of 'interop hacking week' but deeply anchors the next phase of Ethereum's technology roadmap.
Mainline One: Fusaka Upgrade, Initiating a New Cycle of Execution Layer Performance
A major technical focus of this Berlinterop is the Fusaka upgrade. Developers initiated two testnets (fusaka-devnet-1 and berlinterop-devnet-2) and conducted in-depth testing around performance enhancements, block building logic optimization, and parallel path experiments.
Fusaka, as an important upgrade of Ethereum's execution layer, aims not only to improve throughput but may also pave the way for future Pectra upgrades (such as parallel EVM, short slots, etc.).
This development week also clarified the rhythm for the upcoming weeks: Fusaka-devnet-2 is expected to be released after community discussions, and after testing is completed, it will be deployed to the Sepolia testnet by the end of summer and enter the mainnet upgrade preparation stage.
This is Ethereum's most direct response to 'performance' in recent years and is seen as a key signal for opening up to modular and high-performance L2 protocols.
Mainline Two: L2 and Mainnet Collaboration Mechanism Restructuring, Accelerating Interoperability Standardization
Berlinterop specifically set up a dedicated L2 collaboration day, inviting representatives from teams such as Arbitrum, Base, OP Labs, Polygon, Scroll, Starkware, World Chain, ZKsync, etc., to discuss the current situation and bottlenecks of L1 and L2 collaboration.
Consensus is forming: Layer 2 is no longer just a user of Ethereum, but also a co-builder of the protocol and an experimental field for scaling paths.
During the meeting, the L2 team raised three main demands:
As a mainnet user, I hope to gain more data availability resources (such as blobs) and faster final confirmation speeds;
As a protocol stakeholder, I hope to be considered when EVM changes are made and to prepare in advance;
As an operator of a high-throughput execution layer, I hope to participate in the standard-setting for mainnet scaling paths based on past experiences.
This marks a turning point, Layer 2 is no longer just an 'accelerator' for Ethereum; it is becoming a part of the mainnet operational mechanism.
We can foresee that future EIP proposals, gas models, etc., will see more depth of involvement and design collaboration from L2, laying a standardized foundation for the future of 'multi-chain collaboration + cross-chain abstraction.'
Mainline Three: Accelerating towards a Modular Execution World
Another key scene of Berlinterop is the zk session. Researchers from dozens of projects, including Scroll, Succinct, Starkware, ZKsync, ZKM, and RISC Zero, engaged in candid and in-depth discussions around zkEVM, Stateless Client, and ISA standardization.
Among them, the Stateless Client route is particularly focused on — a lightweight client that does not rely on local state and verifies block validity through zk proofs is becoming a core candidate for the next generation of Ethereum nodes.
They plan to deliver prototype implementations by the end of 2025 and address:
Proof Incentive Mechanism (Who generates? Who verifies?)
Design of censorship-resistant data sources
zk VM Instruction Set and Compilation Path Standardization
This work will fundamentally change Ethereum's definition of 'node': lightweight, verifiable, and modular execution will gradually become the default form. It represents: we are one step closer to the future of 'compressing Ethereum into a browser plugin.'
The main storyline of Ethereum has emerged.
Faster execution layer → Fusaka fires the first shot
Stronger collaboration mechanisms → L2 officially takes the stage
Lighter clients → zk and modular execution are being implemented
All of this constitutes the triple upgrade challenges and opportunities that Ethereum will face in 2025.
Written at the end
Berlinterop sends a clear signal to the entire ecosystem:
The pace of Ethereum's technological iteration is accelerating, and the collaborative relationship between Layer 2 and the mainnet has also entered deep waters. Execution performance, interoperability, and zk modularity constitute the three most certain main lines for 2025.
For Layer 2, which is built on Superchain and committed to creating the next generation of on-chain NFT and Agent infrastructure, Mint closely follows and understands future key technological trends. We firmly believe that only by continuously promoting the upgrade of the underlying protocol and ecological collaboration can we bring users a more efficient, secure, and innovative blockchain experience.
Mint is already on the way.
