Ethereum co-founder Vitalik Buterin has highlighted the importance of Peer Data Availability Sampling (PeerDAS) as a strategic solution to the rising blob storage demands on the Ethereum network. This feature is part of the upcoming Fusaka upgrade, timed with Ethereum’s recent achievement of processing six blobs per block, which has raised significant concerns about data bloat across the ecosystem.
Blobs, introduced via Ethereum Improvement Proposal 4844 (EIP-4844), serve as temporary on-chain data containers intended to reduce costs for Layer-2 rollups while minimizing long-term storage burdens. Unlike call data, which remains on the blockchain indefinitely, blobs are designed to expire after about two weeks. This structure not only lowers the expenses associated with operating rollups but also enhances the overall scalability of Ethereum.
As these blobs gain traction, the analysis from on-chain expert Hildobby indicates that several Layer-2 solutions, including Base, Worldcoin, Soneium, and Scroll, have come to rely heavily on them. However, this adoption comes with challenges; validators now require over 70 gigabytes of storage for blobs, a number that could potentially surge to more than 1.2 terabytes if data is not pruned effectively. This significant surge has prompted developers to seek solutions that will maintain both scalability and storage efficiency.
Buterin elaborated on how PeerDAS operates to tackle these issues by distributing data storage responsibilities across the network rather than relying solely on individual nodes. He explained that each node will request only a limited number of data “chunks” to probabilistically verify the availability of over 50% of the total chunks. When that threshold is met, a node can then download and reconstruct the entire dataset using erasure coding.
Despite the promise of PeerDAS, Buterin noted that complete block data will still be required at certain moments, such as during the initial broadcast of a block or if reconstruction from partial data is necessary. To prevent data manipulation, he emphasized the need for “honest actors” within the network. However, he believes that PeerDAS is robust enough to handle scenarios even with considerable dishonesty, as other nodes can step in to fulfill necessary roles.
Nevertheless, Ethereum’s core developers are approaching the rollout of PeerDAS with caution, despite their extensive research on it. To mitigate potential risks, they have decided to implement a phased approach using Blob Parameter Only (BPO) forks rather than an immediate full-scale deployment. The initial fork, set for December 17, aims to increase blob limits from six out of nine to ten out of fifteen. A subsequent fork, scheduled for January 7, 2026, will further raise these limits to fourteen out of twenty-one.
This gradual strategy allows developers to closely monitor network performance and make necessary adjustments. Buterin anticipates that blob counts will increase, setting the stage for more substantial changes in the future. He believes that PeerDAS will be crucial for supporting the growth of Layer-2 solutions and preparing Ethereum’s base layer to accommodate higher gas limits, ultimately facilitating the migration of execution data entirely into blobs.
