Ethereum's Rollup-Centric Roadmap Explained

Ethereum's rollup-centric roadmap is the strategy, laid out by Vitalik Buterin and the core Ethereum research community starting around 2020, of scaling the network primarily by pushing transaction execution off-chain onto Layer 2 rollups, while Ethereum's base layer focuses on being an efficient, secure settlement and data availability layer for those rollups rather than trying to process every transaction itself.

The original plan Ethereum moved away from

Ethereum's early scaling roadmap centered on "sharding" — splitting the network into many parallel chains (shards) that could each execute their own transactions, multiplying overall throughput. This approach required each shard to independently handle execution, which turned out to be an enormously complex engineering problem, particularly around ensuring cross-shard communication stayed secure and efficient.

Around 2020, as rollup technology matured faster than expected, Vitalik Buterin proposed a shift: instead of building execution sharding, Ethereum could get most of the same scaling benefit faster and more safely by leaning on rollups, which were already demonstrating that transactions could be executed off-chain and still inherit Ethereum's security through fraud proofs or validity proofs. This became known as the rollup-centric roadmap.

The division of labor

Under this model, responsibilities split cleanly:

• Ethereum L1 focuses on consensus, settlement, and data availability — being the secure, decentralized foundation that rollups can post their data and proofs to.
• Rollups (L2s) handle actual transaction execution, batching many user transactions together before settling a compressed summary back to L1, using either fraud proofs (optimistic rollups) or validity proofs (ZK-rollups).

This is the same layering discussed in L2 vs L3, extended in some cases even further with Layer 3s building on top of L2s.

Key milestones along the roadmap

Milestone	What it aimed to accomplish
The Merge (2022)	Transitioned Ethereum from proof-of-work to proof-of-stake, a prerequisite for later scaling stages
The Surge	Focused on rollup scaling, including EIP-4844 (proto-danksharding), which sharply cut the cost of posting rollup data
The Verge	Aims to make it easier and cheaper to verify Ethereum blocks, including through advances like Verkle trees
The Purge	Aims to reduce the technical debt and storage burden of running an Ethereum node
The Splurge	Covers a range of smaller improvements that don't fit neatly into the other categories

This "Merge, Surge, Verge, Purge, Splurge" framing is a commonly used shorthand for organizing the roadmap's various stages, though the underlying work spans many individual proposals rather than a strict, sequential checklist.

Why this approach won out over execution sharding

The rollup-centric approach let real scaling benefits arrive years earlier than a from-scratch sharding implementation likely would have, since it built on rollup technology that teams were already actively developing and deploying independently. It also meant the ecosystem could diversify — different rollups could experiment with different trade-offs (optimistic versus ZK, different levels of EVM equivalence) rather than Ethereum committing to one single, monolithic scaling design.

The trade-off is that scaling responsibility, and to some degree security assumptions, are now distributed across many independent rollup teams rather than being handled uniformly by Ethereum's own protocol — meaning the practical security and decentralization of "using Ethereum" today depends partly on which specific L2 you're using, not just on Ethereum's base layer alone.

What this means for users today

Nearly every major low-fee chain you interact with in the Ethereum ecosystem — Arbitrum, Base, Optimism, zkSync Era, Starknet, Polygon zkEVM, Linea, and others — exists because of this roadmap decision. Comparing gas fees on L2 to Ethereum mainnet, or checking why gas fees dropped further after EIP-4844, are really both downstream consequences of this single strategic choice made years ago. You can track how activity is currently distributed across this rollup ecosystem on /chains and /defi.

Why this created a more fragmented, but more experimental, ecosystem

One underappreciated consequence of the rollup-centric approach is that Ethereum's scaling landscape today looks far more fragmented than a single-sharding design would have produced — dozens of separate L2s, each with its own sequencer, bridge, and sometimes its own Layer 3s built on top. That fragmentation brings real costs, particularly around liquidity being spread thin across many networks and users needing to manage multiple bridges and network configurations. But it also allowed rapid experimentation: optimistic and ZK approaches, different zkEVM types, and app-specific rollup appchains could all be tried in parallel, with the market effectively selecting which designs proved durable, rather than the ecosystem waiting on one single protocol-level scaling design to be finished and proven correct before anyone could benefit from it.

Bottom line

Ethereum's rollup-centric roadmap shifted the network's scaling strategy from building execution sharding itself to letting Layer 2 rollups handle execution, while Ethereum focuses on cheap, secure settlement and data availability. That decision, refined through milestones like the Merge and the Surge, is the direct reason today's Ethereum L2 ecosystem — and its comparatively low fees — exists in its current form.