Layer‑2s eating L1 volume
Layer‑2 rollups and modular approaches are continuing to grab transaction volume and capacity from Layer‑1 chains, with posts noting 10x throughput gains and cheaper data-availability via blob solutions like EIP‑4844. Analysts are framing modular stacks as the place where scaling economics are playing out now. (x.com, x.com)
A layer‑2 is a second lane built on top of a base blockchain, and in 2026 those second lanes are handling more of Ethereum’s day‑to‑day traffic than the main chain itself. Ethereum’s own roadmap says rollups are the primary way it plans to support more users while keeping mainnet security. (ethereum.org) Ethereum made that design cheaper on March 13, 2024, when the Dencun upgrade activated Ethereum Improvement Proposal 4844, a change that added temporary “blobs” for rollups to post data at lower cost. Ethereum.org says blobs stay available for about 18 days instead of being stored forever by every node. (blog.ethereum.org, ethereum.org) That matters because rollups used to publish transaction data as call data, which Ethereum processes fully and stores permanently. Ethereum.org says blob transactions were designed to cut that cost and are part of a path to more than 100,000 transactions per second through later danksharding upgrades. (ethereum.org) The result is a market where execution is moving off the base chain while Ethereum keeps settlement and security. L2BEAT’s activity and summary dashboards now track a scaling ecosystem large enough that analysts use it as the main scoreboard for Ethereum transaction throughput, not just a side market. (l2beat.com, l2beat.com) A modular blockchain takes that split further by separating jobs that one chain used to do together: ordering data, executing transactions, and final settlement. Celestia’s documentation describes its network as a data‑availability layer that orders blobs and keeps them downloadable while execution and settlement happen on other layers. (docs.celestia.org) Celestia says light nodes can check whether block data exists by sampling small pieces instead of downloading the whole block. That design lets chains buy data availability as a service, much as rollups buy blockspace from Ethereum blobs. (docs.celestia.org) The software stacks around those chains are multiplying as well. Optimism says the Open Platform Stack is a standardized open‑source framework for production layer‑2 blockchains, and Arbitrum says its Orbit system lets developers launch custom rollup or AnyTrust chains with Nitro. (docs.optimism.io, arbitrum.foundation) That has changed what “scaling Ethereum” means in practice. Instead of one chain getting steadily bigger, the current model is a base layer selling security and data space to many faster chains that compete on fees, app design, and developer tooling. (ethereum.org, docs.optimism.io, arbitrum.foundation) The tradeoff is that not every high‑throughput chain carries the same security assumptions. L2BEAT and related safety trackers separate fully secured rollups from systems that add committees, alternative data availability, or other trust assumptions to reach lower fees. (l2beat.com, l2safety.info) So the current fight is less about whether blockchains can scale than about where each function sits and who captures the fees. Ethereum is positioning mainnet as the settlement layer, while rollups and modular data layers are competing to absorb the volume that used to hit layer‑1 directly. (ethereum.org, docs.celestia.org, l2beat.com)
