EIP-4844 shipped in March 2024 promising cheaper data posting for Layer 2 rollups, and on that narrow metric it delivered. Blob fees collapsed relative to pre-4844 calldata costs, and users on Arbitrum, Base, and Optimism saw transaction fees drop to fractions of a cent almost overnight. What nobody fully modeled was what happens when you successfully decouple L2 settlement costs from mainnet gas - and what that decoupling does to Ethereum’s fee revenue picture.
The Revenue Thinning Nobody Wants to Say Out Loud
Ethereum’s base fee burn, the mechanism that gives ETH its supply-side pressure since EIP-1559, depends heavily on activity and congestion on the execution layer. As L2s have absorbed more user activity - and blob fees remain structurally lower than calldata ever was - less economic activity gets expressed as base fee burn on L1. Ultrasound.money data has shown ETH issuance net of burns has periodically flipped into net-inflationary territory through 2025. That’s not a bug in blob design exactly, but it’s a consequence that the community spent less time stress-testing than the scaling benefits.
The tradeoff isn’t fatal. Ethereum’s staking rewards remain stable, validator participation is near all-time highs above 30 million ETH staked, and the network processes more transactions in aggregate than at any point in its history. But those transactions increasingly live on Layer 2s where Ethereum earns settlement fees, not execution fees - and blob fees are intentionally priced to stay cheap.
Blob Saturation Is Closer Than It Looks
Here’s the part that shifts the story. The current blob capacity - six blobs per block as of Dencun - is not unlimited. During high-activity periods, blob base fees have spiked sharply, temporarily reversing the L2 fee compression. Pectra, which went live on mainnet in May 2025, increased the target and max blob count, providing headroom. But the fundamental dynamic remains: blob demand scales with L2 adoption, and L2 adoption is accelerating faster than most base-layer capacity estimates assumed.
Ethereum’s roadmap response is Fusaka, slated to include PeerDAS - a more sophisticated data availability scheme that distributes blob data across the validator set rather than requiring all nodes to store all blobs. If PeerDAS ships cleanly, the blob ceiling rises substantially without proportional node storage costs. That’s the theory.
Whether it resolves cleanly enough to keep pace with L2 growth before blob markets tighten again is the question that doesn’t have a comfortable answer yet.