Introducing it in June protocolto reorganize the Ethereum Foundation’s research and development team to better fit current strategic objectives; Scale L1, Scale bloband Improve UX Without compromising Ethereum’s commitment to safety and hardness.
Over the next few weeks, we will publish updates for each workstream, covering ongoing progress, new initiatives, open questions and collaboration opportunities. Today we’ll start with scale L1 —-hoping for follow-up on scale blobs and we’ll improve our UX soon!
tl; dr
- Marius van der Widdin Join Ansgar Dietrichs and Tim Beiko for L1
- Mainnet gas limit increased after 45m –berlininterop100m of gas and the first step on the road to more
- All major execution layer clients have been shipped Your registration history expiration date in advancesignificantly reduces node disk usage
- Block-Level Access List (BALS) is considered the headliner for Gram Stardom
- Computing and State Benchmarking initiatives are underway to better manage EVM resource pricing and performance bottlenecks
- The road to Zkevm real-time evidence is becoming more specificZK-based Attester client prototyping is underway
- We’re still hiring Performance Engineering Lead: Application will be closed on August 10th
I’m serious about L1 scaling
Scaling Ethereum requires tailoring ambitious design and engineering pragmatism. To achieve this, we have appointed Marius van der Widdin Together with Ansgar Dietrichs and Tim Beiko, he is co-leading in Scale L1.
Marius’ extensive engineering experience with Geth and his commitment to protocol security, he is perfectly suited to align his scaling strategy with Ethereum’s constraints.
Together, Ansugar, Marius and Tim defined a set of important initiatives that will help L1 scale as quickly as possible.
Towards the 100m main net gas limit
The immediate goal is to safely scale Ethereum’s mainnet gas limit to 100m per block. Parithosh Jayanthibeing closely supported by Nethermind’s Perfnet team, leads our work Each increase increases.
recently berlinterop eventThe client team has significantly improved its worst performance benchmark, allowing for the recent increase to 45m gas.
Additionally, client curing has become an integral part of the 100m gas initiative. The Pectra upgrade rollout highlighted some issues caused by network instability. Even if the network temporarily loses finality, it is paramount to ensure that clients remain robust as throughput increases.
History expiry date
History expiration project led by Matt Garnettreduces historical data footprint for Ethereum nodes. Recent developments of Partial history expiration date I deleted the merge historical data in advance and saved about 300-500 GB of disk space on the full node. This allows you to run comfortably on a 2TB disk.
Based on this, we are currently developing a rolling history expiration date. This makes the storage needs of the node easier to manage with Ethereum scaling.
Block-level access list
Block-Level Access List (BALS) is defended Toni Worstrutterhas emerged as a major candidate for inclusion in Gram Stardom upgrades. Vals offers several important benefits:
- Enables parallel transaction execution within a block.
- It facilitates parallel computation of state roots and greatly accelerates block processing.
- It allows preloading of required states at the start of block execution and optimizes disk access patterns.
- Improves overall node synchronization efficiency and benefits new and archive nodes.
These improvements collectively enhance Ethereum’s ability to ensure higher gas limits and faster blocking.
Benchmarking and pricing
The ongoing challenges at Scaling Ethereum match the gas costs of EVM operations to the calculation overhead. Worst Edge Case Performance is currently limiting network throughput.
By improving benchmark infrastructure and recogging operations that clients cannot optimize, block execution times can be more consistent. If you close the gap between the worst case block and the average block, you can raise the gas limit at the same time.
Ansgar Dietrichs is directly informed by Perfnet’s comprehensive benchmarks and leads efforts focused on targeted benchmarks and engineering interventions to identify and resolve computing-heavy bottlenecks. There have already been major advances since Berlittez, especially in managing worst computational scenarios.
in parallel, Carlo Superez Spearhead bloatnet: An initiative aimed at benchmarking and optimizing state performance. This involves testing node performance under conditions where state size is twice as large as the current mainnet, and gas limits reaching 100-150m directly inform both relocation and client optimization.
Both of these efforts will homogenize resource costs across operations to Gramasterdam’s EIP proposal, allowing for L1 scaling.
Zkevm Attester Client
Today, Ethereum nodes perform all transactions within the block when they receive a block. This is computationally expensive. To reduce this computational cost, Ethereum clients can instead verify the ZK proof of the block’s execution. To enable this, you must generate a block proof in real time. Closer.
Kevaundray Wedderburn The main task of the ZKEVM Attester client is to have real-time proofs and use them to fulfill their validator duties.
Once the prototype mainnet is ready, it will be deployed as an optional verification mechanism. We hope that a small group of nodes will adopt this next year, building confidence in its robustness and security.
After this, the Ethereum nodes can gradually migrate to ZK-based validation, which ultimately becomes the default. At that point, the gas limit on L1 could increase significantly — Beast Mode!
RPC Performance and Employment
As throughput increases, different node types (execution, consensus, RPC) face clear challenges. RPC nodes are particularly pressured as they respond to a wide range of historical and real-time state requests.
Internally, EF’s Geth and Pandaops teams are actively studying the optimal configuration of various node types. The importance of this will increase over the next few years and we hope to increase the expertise in this domain.
That’s why we are actively hiring Performance Engineering Lead. The application closes August 10th. If you’re as excited about the scaling of the L1 as we do, we’d love to hear from you!
