Live
USD · 24h
Ethereum & Altcoins

PeerDAS and the scaling of Ethereum data availability

PeerDAS scales Ethereum data throughput by splitting blobs into 128 smaller columns, allowing nodes to verify data through random sampling. This technology enables the network to increase blob capacity from 6 to 64 per block while maintaining manageable hardware requirements for validators.

PeerDAS and the scaling of Ethereum data availability

PeerDAS scales data throughput

PeerDAS provides the most effective path to scale data availability without increasing hardware requirements for home validators. The Fusaka update in December 2025 introduced this technology. PeerDAS splits blobs into 128 smaller pieces called columns or cells. A regular node no longer downloads the entire blob. Instead, a validating node might only be responsible for custodying 8 or more of these 128 columns. This is about 1/8th of the data. Each blob holds 128 KB of data, consisting of 4,096 field elements of 32 bytes each. Erasure coding using Reed-Solomon math allows a node to reconstruct the full blob from just 50% of its original pieces. This system works because nodes randomly sample the network to confirm availability. If they find the pieces they request, they gain high statistical confidence that the data exists. Nodes that consistently fail to provide data get blacklisted by other nodes. A failure probability as low as 1 in $10^{20}$ to $10^{24}$ exists for these sampling methods. Because the transition from blob proofs to cell proofs helps keep hardware requirements manageable for individual stakers, the network can scale without pricing out users who run their own nodes in the Ethereum ecosystem.

Parameter Value/Status
PeerDAS columns per blob 128
Erasure coding threshold 50%
BPO1 Target (Dec 2025) 10 blobs
BPO2 Target (Jan 2026) 14 blobs
Blob pruning period 18 days

Verification through KZG commitments

KZG polynomial commitments allow Ethereum to verify data without the EVM accessing the blobs directly. Rollups post a commitment that fits a polynomial function to the data. Provers check this by evaluating the polynomial at secret points. If the values match, the data is correct. This method slashed rollup fees by over 90 percent after the Dencun upgrade in March 2024. Optimistic rollups need this data to remain available for the challenge period so they can submit fraud proofs. If a sequencer withholds data, the security of the rollup fails. ZK rollups rely on validity proofs to ensure correctness, but users still need data access to reconstruct balances or exit. You should remember that blobs stay on consensus nodes for roughly 18 days before pruning.

The KZG ceremony involved over 140,000 contributions to generate the secret random string used to verify data. Because the commitment evaluates the polynomial at specific points defined by the random numbers in the KZG ceremony, it prevents attackers from generating a new polynomial that fits those points. Rollup security depends on the availability of the data used to produce state changes. Optimistic rollups assume all transactions are valid by default. They use a challenge period to allow operators to submit fraud proofs. If a fraud proof succeeds, the rollup re-executes the disputed transaction batch and reverts to a previous valid state. The operator who included the incorrect transaction forfeits a staked deposit. ZK rollups do not use a challenge period because they provide cryptographic guarantees for every batch. This provides faster finality than optimistic rollups.

Full Danksharding targets

Full Danksharding remains the final goal for scaling Ethereum to millions of transactions per second. This target expands blobs from 6 per block to 64 per block. PeerDAS provides the first step toward this goal. The network already saw blob capacity increases through Blob Parameter Only forks in early 2026. BPO1 in December 2025 increased the target to 10 blobs. BPO2 in January 2026 increased the target to 14 blobs. These forks allow the network to adapt to changing layer 2 needs without waiting for a full hard fork.

The roadmap involves moving from 6 blobs to 64 blobs and eventually much more. Once developers collect data from the BPO upgrades, they will proceed with more aggressive scaling. This includes BPO3 and BPO4 with the aim of reaching 128 blobs per block. The current capacity for blobs follows a schedule that allows for independent parameter changes. Will the network successfully transition to 128 blobs per block without compromising decentralization?

Join the discussion

Leave a Reply

Your email address will not be published. Required fields are marked *