Thanks to Justin for the inspiration.

The Ethereum community has been in uncertainty about how to move forward for a while because of value accrual, price action and management related concerns. We believe that all these natural uncertainties comes from adapting to the new mindset shift of Ethereum as a data availability layer. This is a transition period. We know that as the DA layer matures, Ethereum will have a higher ROI than ever before.

Ethereum moved to a rollup-centric roadmap 5 years ago, where much of the transaction execution is offloaded to rollups, while Ethereum's L1 focuses on providing security, decentralization, and data availability. Thanks to this vision, we've removed the execution bottleneck and scaled it to 15-25x in average usage.

However, due to this execution shift from L1 to L2s, Ethereum started to experience some value accrual problems. Because previously, execution fees (congestion and contention fees) were Ethereum's main source of revenue, but Ethereum doesn't have that anymore.

Blockchains have 3 direct revenue streams; congestion fees (base fee), contention fees (priority fee) and data availability.

In Ethereum, congestion fees were a strong bottleneck during the 2021 bull season. Due to the strong restrictions on lower hardware and bandwidth requirements for decentralization, the gas limit was always low. However, today we have battle-tested and open-sourced fraud-proof and SNARK games such as Arbitrum's BoLD, Optimism's Bedrock, zkSync's Boojum and Starknet's Stwo etc. to solve this computational bottlenecks to 100x by rollups. Today, congestion fees are not even a problem, as the bottleneck has shifted to data availability, which we will cover later.

Contention fees (aka MEV) are impossible to destroy, not as a financial problem, but as a law of thermodynamics. This is why “distribution” is the key solution here, you can not destroy it but you can distribute it. It’s quite normal to expect that MEV will be recaptured by apps, wallet and users gradually in future. The shining of MEV-protected DEXs and app-specific sequencing are a significant signs of this.

From Ethereum’s point of view, congestion and contention fees are doomed to die, and they can’t be a revenue stream for blockchains anymore for the reasons outlined above. Thus we’ve reached to an important one-way crossroad, and this lead to data availability.

Ethereum's revenue model is undergoing a fundamental shift. While historically based on execution fees, Ethereum is transforming into a specialized data availability (DA) layer within its rollup-centric ecosystem.

Post-Merge and EIP-1559 introduced fee burning, connecting ETH's value directly to network usage. However, as L2 rollups handle increasing transaction volumes, Ethereum's L1 is repositioning as a settlement and data availability provider rather than competing with L2s for execution fees. This shift means Ethereum now prioritizes scaling blob capacity — allocating space for rollup data — rather than maximizing execution revenue. While DA services generate less revenue per unit than complex execution tasks, Ethereum's value proposition comes from securing these services at scale (induced demand).

Critics argue this repositioning surrenders lucrative execution fees to L2s and may leave Ethereum vulnerable to competition from specialized DA platforms offering lower costs. However, Ethereum maintains significant advantages through its security guarantees, decentralization, and network effects. As rollup adoption grows, blob fees are expected to increase through induced demand, potentially generating substantial revenue when combined with expanded capacity from technologies like PeerDAS and full Danksharding. Ethereum's long-term dominance rests not on maximizing short-term fees but on becoming the foundation for a scalable, most secure ecosystem.

Ethereum has 3 blobs per slot (~384 KB). With the Pectra upgrade, which is delayed until April, it's planned to double this to 6 blobs per slot (~0.76 MB). The next step stepping stone is bringing PeerDAS in Fusaka.

PeerDAS

The execution bottleneck shifted to data availability, as all node validators are responsible for downloading the entire blob data. For this reason, Ethereum's bandwidth is low to allow home-stakers to download all data synchronously without any delay.

In PeerDAS, each blob is expanded using 1-dimensional erasure coding, transforming it into 128 columns. Instead of requiring each validator to download the entire blob, PeerDAS partitions the responsibility so that nodes store only a fraction (about 1/8) of the total data while maintaining connections to peers that together cover the entire dataset. They request and verify small portions (e.g., 8 cells or columns) using cryptographic proofs (such as KZG polynomial commitments). If enough nodes successfully sample their portions, it is statistically guaranteed that the entire dataset is available across the network.

Blob data can also be reconstructed by collecting 50% of the samples from different validators as needed. Once 50% of the elements are collected and all are valid, the reconstruction is complete.

PeerDAS allows Ethereum to increase blob capacity 4-8x (from 6 to 48 target blobs per block) while maintaining the same bandwidth requirements for validators, as an important stepping stone between Proto-danksharding (EIP-4844) and full Danksharding.

Today we are in the partial outsourcing zone on Ethereum. We're experiencing the possibilities and capabilities of the DA. With upcoming improvements on blob count and size in 3-4 years, Ethereum will have at least 10 million TPS (L1 10 TPS * Rollups 100x * Danksharding 100x * Nielsen's Law 100x) collectively. Thousands of rollups will enjoy Ethereum DA by achieving synchronous state access with the L1 and other rollups. Blob fees will pay much more than the cost of running all boxes in the network at full outsourcing zone.

Ultrasound Rollups

This is the ideal scenario in rollups for Ethereum: every rollup is based. Based rollups will infinitely scale Ethereum to exponential sizes by leveraging Ethereum's sequencing and data engines. They will continue to consume more DA than any other rollup and reach synchronous state access more often.

Reasonable Economics

Ethereum is the network with the most cryptoeconomic security, the most activity, and the most network effect. Ethereum should take these features into account when pricing the DA service it provides. However, it should also take care of reasonable pricing, otherwise rollups might start thinking about moving to other DA layers. See here for the proposal to increase the minimum base fee per blob of gas to speed up price discovery on blob space by Resnick et al.

Get the latest from Taiko:

• Website: https://taiko.xyz.

• Discord: https://discord.gg/taikoxyz.

• GitHub: https://github.com/taikoxyz.

• Twitter: https://twitter.com/taikoxyz.

• Community forum: https://community.taiko.xyz.

• Youtube: https://www.youtube.com/@taikoxyz.

• Warpcast: https://warpcast.com/taikoxyz.

Contribute to Taiko on GitHub and earn a GitPOAP! You will also be featured as a contributor on our README. Get started with the contributing manual.