Ethereum L2 decentralization progress falls short of expectations! Vitalik: The old roadmap is outdated; a new path must be found.

Vitalik proposes revisions to Ethereum’s scaling roadmap, believing that the mainnet performance has improved and that Layer2 should no longer serve as “branded sharding,” but instead focus on diverse functionalities and specialized value.

Ethereum Scaling Blueprint Undergoes Major Revision: Vitalik Suggests Ending the “Branded Sharding” Vision

Ethereum co-founder Vitalik Buterin shared his views yesterday (2/3), sparking a industry-wide reassessment of the role of Layer2 scaling solutions. Vitalik pointed out that the original “Rollup-centric” roadmap is no longer fully applicable in the current context and that a new development path must be sought.

Image source: X/@VitalikButerin Vitalik notes that Ethereum’s original “Rollup-centric” roadmap is no longer fully suitable in the current environment

For a long time, Ethereum developers have pursued scaling goals, aiming to create enough block space to accommodate various applications and transactions without sacrificing security and decentralization. Historically, the Ethereum community generally believed that the mainnet should shift toward handling settlement, transferring most transaction and application loads to third-party built Layer2 scaling networks such as Arbitrum, Optimism, and Base. These networks were once defined as Ethereum’s “Branded Shards,” bearing the social status and responsibility as extensions of the mainnet.

However, Vitalik believes this vision has become disconnected from reality, primarily because Layer2’s progress toward decentralization has fallen short of expectations, and Ethereum’s mainnet (Layer1) has significantly improved its scaling capacity.

Mainnet scaling performance has greatly improved, shifting Layer2’s core value spectrum

Ethereum’s mainnet scaling speed is exceeding expectations, which is a key fact driving the change in perspective. Currently, transaction fees on Ethereum Layer1 remain low, and the Gas Limit is expected to increase substantially in 2026 and in the coming years. As the mainnet continues to scale, Layer1 has become capable of operating independently and no longer relies on Layer2 to serve as “branded sharding.”

Although Layer2 was previously likened to a “overflow room” in a conference hall to reduce congestion in the main venue, advances in technologies like PeerDAS and improvements in data availability have enabled the mainnet to handle data demands more effectively. Vitalik proposes defining Layer2 as a “Full Spectrum” network series. In this spectrum, different networks should offer varying levels of connectivity and technical features based on user needs. Some networks can still be secured by Ethereum’s strong security guarantees, while others can opt for looser connection methods, with users choosing based on personal or automated requirements.

Decentralization bottleneck: Vitalik warns that multi-sig cross-chain does not equal Ethereum scaling

Beyond mainnet performance improvements, Layer2 faces difficulties in progressing toward full decentralization, which is also a reason for Vitalik’s change of stance. According to his previous decentralization architecture proposal, Layer2 networks should evolve from “Stage 0” to “Stage 2,” representing fully trustless, permissionless systems. “Stage 1” is defined as a state with “limited training wheels,” including a decentralized governance security council and permissionless fault proofs. Currently, most Layer2 solutions are still in early stages, with only a few, like Coinbase-supported Base network, successfully reaching Stage 1 last year.

Vitalik openly states that many Layer2 networks cannot meet the security and decentralization standards required for true branded sharding. He is especially critical of networks relying on “multisig bridges.” A network claiming to support 10,000 transactions per second (10k TPS) via an EVM-compatible chain, if connected to Ethereum mainnet through a multisig-controlled cross-chain bridge, does not constitute real “scaling Ethereum.” Genuine scaling must be guaranteed by Ethereum’s trust model, ensuring transactions are valid, censorship-resistant, and irreversible.

Seeking specialized industry competitiveness: Layer2 developers must create unique value propositions

In response to the dramatic changes in the Ethereum ecosystem, Vitalik offers new strategic advice to Layer2 developers: seek value-added services beyond “scalability.” As the mainnet becomes more efficient, simply pursuing lower transaction costs is no longer a sufficient competitive advantage.

Some Layer2 projects, possibly due to regulatory requirements, may choose to remain at Stage 1 indefinitely to maintain ultimate control over customer needs.

Vitalik believes this is appropriate for certain clients, but such systems should not be described as scaling Ethereum. He encourages Layer2 teams to focus on “extreme performance” or “specialized architectures” that are difficult to implement on the mainnet, such as non-EVM virtual machines designed for specific applications, privacy features, or ultra-low latency ordering properties. For existing Layer2 projects, Vitalik’s bottom line is that they must at least reach Stage 1 decentralization, especially for networks involving Ethereum (ETH) assets. The value of Layer2 lies in “creating new things,” not merely extending the mainnet. This revision of the roadmap is a pragmatic adjustment under current technological conditions and a realistic recalibration of Ethereum’s multi-layer vision.