Vitalik Buterin announced a shift in Ethereum’s scaling strategy on Tuesday. The Ethereum co-founder said the original rollup-centric roadmap needs replacement as Layer 2 networks failed to decentralize as planned.
Buterin posted on X that L2s decentralized “far slower” than expected while the Ethereum base layer made unexpected progress. This marks a reversal from his previous vision where branded L2 shards would handle most transactions.
The rollup-centric approach launched with the idea that L2s would process transactions at high speed and low cost. These networks were supposed to inherit Ethereum’s security guarantees while operating more efficiently.
However, very few L2s have reached Stage 1 decentralization. Base, the Coinbase-backed L2, advanced to Stage 1 in 2025 by launching permissionless fault proofs and decentralizing its security council.
Most other major L2s including Arbitrum, Optimism and Starknet remain at earlier stages. Buterin previously created a framework measuring rollup decentralization from Stage 0 to Stage 2, with Stage 2 representing full trustlessness.
Buterin identified both technical challenges and regulatory concerns slowing L2 decentralization. Some L2 teams stated they may never progress beyond Stage 1 due to customer regulatory requirements.
Corporate interests also played a role in delayed decentralization. One team explicitly said their customers need them to maintain ultimate control over the network.
Buterin criticized L2s using multisig bridges for Ethereum connection. He stated these networks are not scaling Ethereum but rather operating as competing L1 chains with bridges.
The Ethereum mainnet scaled faster than anticipated through several improvements. Developers raised the gas limit from 60 million to 80 million in January 2026 following a hard fork.
This increase allows more transactions and smart contract operations per block. The change boosts throughput while potentially lowering transaction fees.
Buterin expressed growing interest in native rollups that are built into Ethereum itself. These rollups would be verified directly by Ethereum validators rather than operating separately.
The integration of zkEVM proofs into the base layer would enable native rollups to scale the mainnet. These precompiled features would auto-upgrade with Ethereum protocol changes.
Buterin said L2s should still reach Stage 1 decentralization at minimum. Beyond that requirement, he wants L2s to focus on specialized features rather than pure scaling.
Suggested specializations include privacy features, app-specific architectures, non-financial use cases and AI applications. L2s could also target ultra-low-latency or unique sequencing properties.
The pivot received support from Ethereum community members. Ryan Sean Adams, co-host of Bankless, endorsed the direction with a focus on strengthening the base layer.
Ethereum researcher Justin Drake previously outlined plans for 10,000 transactions per second on the mainnet. The network currently processes between 15 and 30 transactions per second.
Max Resnick, a former Consensys researcher, had advocated for mainnet scaling before moving to Solana. His push for L1 improvements failed to gain sufficient support at the time.
Buterin clarified that L2s still have a role in the Ethereum ecosystem. He wants teams to view L2s as a spectrum of options with varying connections to Ethereum rather than branded shards.
The post Ethereum’s Big Pivot: Why Vitalik Buterin Is Moving Away from Layer 2 Rollups appeared first on CoinCentral.


