PeerDAS expands Ethereum blob capacity and L1 throughput

Developers scheduled the Fusaka upgrade for December 3 after the third and final testnet cleared its dress rehearsal this week. This update introduces PeerDAS, a new networking protocol. PeerDAS allows nodes to verify blob data availability through sampling instead of downloading entire blobs. Erasure coding allows nodes to sample portions of blob data while providing a cryptographic guarantee that the full data remains available across the network. Because the PeerDAS mechanism allows nodes to verify data availability through sampling small portions, Ethereum can increase blob throughput without significantly increasing the bandwidth or storage requirements for individual nodes on the network. The current limit of 6 blobs per block can scale towards 128. This capacity increase helps rollups like Base, Arbitrum, and Lighter manage more transactions. These rollups might see transaction costs drop by 40% to 60% as data availability costs decrease. The 2024 Dencun upgrade introduced blobs to allow for the temporary storage of layer-2 data on Ethereum transaction blocks. This innovation reduced both gas fees and processing time for layer-2 transactions. Fusaka expands on that innovation by increasing the blob space available in each Ethereum transaction block. Vitalik Buterin calls PeerDAS the way to scale layer-2 networks. This change supports the long-term goal of making Ethereum a global settlement layer for all on-chain transactions. Fusaka also introduces a lower bound on the blob base fee through EIP 7918. This ties the blob fee to a fraction of the L1 base fee, specifically the execution base fee divided by 16. This mechanism prevents blob prices from collapsing to near zero during quiet periods and creates a more consistent revenue and burn stream as activity migrates to rollups.

L1 gas limits and security updates

The protocol increases L1 execution capacity through EIP-7935. This raises the default gas limit to 60M, which provides more headroom during demand spikes. EIP-7825 implements spam resistance checks to protect nodes from excessive gas consumption. This proposal sets a protocol-level transaction gas limit cap of 16,777,216 gas. The cap prevents individual transactions from consuming the entire block and prepares the network for parallel transaction processing. Additionally, EIP-7883 and EIP-7823 optimize the ModExp precompile. EIP-7883 increases gas costs to reflect computational complexity, including raising minimal gas costs and tripling general cost calculations. EIP-7823 sets upper bounds for ModExp operations. These changes ensure that resource-intensive cryptographic operations remain properly priced. The increase in gas limit supports the current low fee landscape where fees stay below 0.4 gwei. This expansion enables more computation per block and allows for more complex contract execution.

Component Specific Value or Detail
Default Gas Limit (EIP-7935) 60,000,000
Transaction Gas Cap (EIP-7825) 16,777,216
PeerDAS Storage Requirement ~1/8 of blob data
BPO1 Blob Target / Max 10 / 15
BPO2 Blob Target / Max 14 / 21
L2 Fee Reduction Estimate 40% to 60%

Hardware integration and efficiency

User experience improvements include native support for the secp256r1 elliptic curve via EIP-7951. This adds support for hardware like the Apple Secure Enclave, Android Keystore, and FIDO2/WebAuthn devices. Using these familiar authentication flows reduces friction for mainstream users. EIP-7939 introduces the CLZ opcode to provide a gas-efficient way to perform bit-counting operations. This addition helps mathematical operations, compression algorithms, and post-quantum signature schemes. It also reduces ZK proving costs. To manage the expansion of blob space, the network uses Blob Parameter Only (BPO) forks. These minimal, config-only upgrades allow developers to adjust blob targets without a full hard fork. BPO1 increases the per-block blob target to 10 and the maximum to 15. BPO2 follows to increase these to 14 and 21. The first BPO adjustment begins on December 9, 2025, while the second occurs on January 7, 2026. You should check if your specific wallet provider supports these new cryptographic standards.

Testing reveals unexpected network behaviors

The transition to PeerDAS revealed unexpected behaviors in the networking layer. Around epoch 411440, the attestation participation rate fell from 97% to roughly 77%. This decline matched the estimated share of Prysm validators. Stale attestations caused Prysm nodes to regenerate old states instead of operating on the most recent head. This behavior related to a heuristic from the Capella era. Prysm operators applied a workaround by adding a flag to their beacon node to disable the generation of old states. Will these unexpected behaviors in fork-choice rules impact the upcoming Glamsterdam upgrade? Additionally, the Dasmon probing tool reported missing data due to late blob deliveries. These reports likely stem from late arrivals rather than structural failures. A Nethermind-Nimbus validation mismatch also occurred at slot 13164552. This issue appeared when Nethermind ran in a specific Nix-based environment and affected a very small number of users. Furthermore, a block at slot 13164887 arrived 4.65 seconds late. This delay created a scenario where a late-arriving block became part of the canonical chain after receiving enough attestations. The development team replaces the retiring Holesky testnet with Hoodi, Sepolia, and Ephemery. Hoodi serves as the primary environment for validator and staking provider testing, while Sepolia remains the staging ground for dApp developers. Ephemery provides quick-reset validator cycles every 28 days for specific testing scenarios.

Newsletter