Protocol update 002-scale blove

Follow -up Protocol update 001We want to introduce a way of approach to Blob Scaling. L1 plays a powerful basis for the L2 system, expands Ethereum and requires the required components of the Secure L2 solution is the data solubility provided by L1. Anyone can verify the update that L2 returns to L1 through data availability. Blobs is a data solubility unit of the protocol today, so scaling the Blob Count per block is a major requirement for adoption of L2 for real -time payment, defi, social media, games and AI/agent applications.

Our work consists of a series of gradual changes in Ethereum’s Blob Architecture. To accelerate our scaling speed, we expand from the “fork -centered” philosophy and send gradual optimization in an unbroken manner. Therefore, we are related to the two network upgrades, with the project related to the following projects and the period (“Interfork”) between them.

Tl; doctor

• Anger is introduced BloomBeyond today’s throughput level, to 6 blobs/blocks up to 48 blobs/blocks, new data architecture that allows blobs to be blended scaling
• BLOB parameter only (BPO) fork Gradually, the Mainnet Blob Count is strengthened with the gradual peer -to -peer bandwidth optimization.
• The advanced networking technology planned for Glamsterdam will expand further by repeating peerdas design.
• Mempool sharding preserves Ether Leeum’s value As the data continues to expand
• Research on next -generation DAS unlocks the evolution of safe DA scaling.

Fsuka peerdas

The first milestone is to provide peerdas in the upgrade of the upgrade of the upgrade. Peerdas introduces DAS (Data Availability Sampling), where individual nodes download only sub -sets of blob data from a given block. Along with random sampling per node, even if the total number of blobs increases, the calculation load is limited. The node no longer needs to download all idiots from the block, so you can increase the number of blobs without significantly increasing node requirements.

Fusaka is expected to implement all Ethereum customers later this year. A wide range of tests were performed in the development network (“Devnets”), including non -assigned scenarios and “Data withholding” conditions. At this point of the R & D process, we are planning to continue to strengthen existing denots and place them on Testnets and Mainnet. Barnabas The final stage of the upgrade pipeline is leading the charging to ensure smooth progress.

Peerdas v1.x

We have two branches in the strategy to gradually expand the blobs between BPOS and bandwidth optimization between Fusaka and GlamsterDam upgrades. They are additional because they can use resources for higher throughput by using better bandwidth.

BPO

The peerdas introduced in fusaka sets the steps for 8 times the theoretical increase in the throughput of Ether Leeum (ie, to 64KB/s ~ 512kb/s) today. At the time of the fusaka deployment, the core developers decided to gradually increase instead of jumping immediately. “Only BLOB parameters” hard fork. Through this mechanism, core developers can continue to grow the growth trajectory by programming the automatic increase in blobs over time. BPOs do not require manual interventions to activate when programmed, and multiple units of BPO steps may be included in the same client release. During the stage, we open the way for the next increase in response to the scaling bottleneck phenomenon that can be monitored and presented only to the main net. Barnabas Busa, along with the other people of the EF Pandaops team, tightly cooperates with the customer team to distill the right schedule to achieve 8 times scaling from today.

Bandwidth optimization

There are many things you can do to use bandwidth more efficiently in the network. Raúl Kripalani together Munizaga Marco This is a major effort to work on this network engineering. In particular, promising optimization is to introduce “cell level messaging.” This can query more intelligently about some of the samples introduced in Pierdas. These changes can reduce overlapping communication in the network, and bandwidth can be dedicated to safe provisioning with much more blobs. There is no need to change agreement or execution protocol to unlock this milestone, so you can deliver Interfork before next year.

Peerdas v2

This project is the next-generation peerdas design that provides more scale while utilizing the bandwidth saving realized in EIP-7732 (to be included in the Glamsterdam). There is an additional improvement of cell level messaging and data reconstruction technology. The node allows you to sample nodes more flexibly so that you can express the core ideas of DAS as a whole. These profits are set to expand beyond the limits of the imminent peerdas design, along with the benefits of pipelines that can use the time between the blocks more efficiently. There are many moving sculptures, and since the Blob Count is actually expanded in the production environment, the exact number must be corrected to both implementation performance and main net analysis, but this work should provide the final drainage of the DA throughput before finding the alternative design.

This update deployment enters the GlamsterDam upgrade, which is expected in mid -2026. Alex Stokes And Raúl KRIPALANI adjusts R & D here to continue to expand Blob throughput.

Blobpool scaling

The advantage of scaling is clear, but you have to do so while preserving the core value of Ether Leeum. One of these directly related to Blob Scaling is censorship resistance. Mempool serves as a distributed network for including BLOB, and provides censorship resistance directly in the centralized architect network that produces most blocks in Ether Lee. Censorship cases have been improved over time, but unlike the scaling strategy, there is also a blove member scale.

Csaba king We are leading the work here to maintain this important resource. The current implementation supports short -term blove throughput. vigorous research It is the best way to expand the mempool while reaching a higher level of unlocked fusaka.

DA’s future

Beyond the repetition of peerdas, we have a variety of research directions that continue to expand the DA while maintaining the security characteristics of Ethereum. Proposal is generally belonged to Moni Kerr Fullas There are many flavors of active investigation. The core elements of this proposal can all be resistant to ineligible actors while giving up more and more samples, including innovation of peer -to -peer networking. The following work Powerful distributed array It formulates this concept. Other considerations include low intensity, censorship resistance, and evolution of the BLOB fee market, making it easy to obtain Blobs Onchain.

Here research is managed FranceSco D ‘Amato And very active -if you want to collaborate, please contact us!