Ethereum co-founder Vitalik Buterin believes Peer Data Availability Sampling (PeerDAS), a core feature of the upcoming Fusaka upgrade, will fix Layer-2 scalability issues for the Ethereum blockchain. Speaking Thursday in response to an X thread by Dragon XYZ’s head of data, Hildebert Moulié, Buterin explained that the technology will help nodes verify data without requiring them to download the entire blockchain. “The way PeerDAS works is that each node only asks for a small number of chunks, as a way of probabilistically verifying that more than 50% of chunks are available,” Buterin wrote on X. “If more than 50% of chunks are available, then the node theoretically can download those chunks, and use erasure coding to recover the rest.” The programmer reiterated that this would allow a live blockchain on Ether to operate without any single node having to store full datasets. Ethereum network reaches blobs per block record Late Wednesday, Moulié noted that Ethereum had reached its target of six blobs per block for the first time in history. Blobs, or Binary Large Objects, were introduced to Ethereum through the Dencun upgrade in March 2024 under EIP-4844 , also known as proto-danksharding. According to Moulié, blob usage during the spike was largely driven by activity across several L2 rollups. Base accounted for 42% of blob activity, followed by World with 25%, while other contributors included Arbitrum at 8%, OP Mainnet at 4%, and Soneium and Scroll at 3% each. 5/ current used blobspace % 42% Base 25% World 8% Arbitrum 4% OP Mainnet 3% Soneium 3% Scroll 2% Unichain 2% Shape 2% Lighter 1% Ink pic.twitter.com/J6EWWp0VF7 — hildobby (@hildobby) September 24, 2025 Blobs were developed to reduce transaction costs for Layer-2 networks by providing dedicated storage for compressed data. They may have proven effective before, but per Moulié, forecasting blob usage is difficult due to “inconsistent posting by different chains.” The Dragon XYZ head of data also mentioned the current requirements for validator storage exceed 70 gigabytes of space for uncompressed blobs, with estimates topping 1,200 gigabytes for unpruned data. Each blob carries compressed information, but compression ratios vary by chain and rarely surpass 25%. “Each stack has its own compression algo, but generally the more diverse the chain’s usage is, the harder it is to reach higher compression ratios,” he remarked. According to Moulié, more than 10% of blobs are not fully utilized, originating from smaller rollups that post data regularly before blobs reach their 128-kilobyte limit. He also said around 30% of blocks containing blobs have been affected by miner extractable value (MEV) issues. Fusaka will allow different nodes to perform tasks for different blocks EF co-founder Buterin addressed some of the L2 scalability problems, saying that in the initial PeerDAS launch, there will still be cases requiring full data storage. “In the first version, there are two cases where the full data of a block still needs to exist in one place: initial broadcasting, and reconstruction in case a publisher only provides between 50% and 100% of a block,” he surmised. He clarified that the protocol requires only a single honest actor to uphold its integrity, even if several others attempt to act dishonestly. “We only need one honest actor to do them; if there are also 100 dishonest actors, the protocol simply bypasses them. And different nodes can perform this task for different blocks,” Buterin noted. Fusaka upgrade timeline slated for December As reported by Cryptopolitan, the Fusaka upgrade , scheduled for December 3, will formally introduce EIP-7594 and double blob capacity from the current target and maximum of six and nine blobs per block. Ether developers are planning a phased approach like the previous upgrades, with two Blob Parameter Only forks to gradually increase capacity. The maximum blob count will be increased from nine to fifteen in the first fork, then from sixteen to twenty-one in the second, according to Ethereum researcher Christine Kim. The second fork is scheduled for January. Although PeerDAS and related methods have been developed for a long time, Buterin warned developers that they are still untested. Get seen where it counts. Advertise in Cryptopolitan Research and reach crypto’s sharpest investors and builders.
