Skip to content

Deep dive · Mechanisms & Tech · 10 min

Ethereum’s Roadmap: from the Merge to the Verge

The Surge, the Scourge, the Verge, the Purge, the Splurge — a plain-English tour of where Ethereum is actually going next.

On this page

Sources

Why this matters

Ethereum is not a static protocol. It is the most ambitious long-running software upgrade project in crypto, with a roadmap that has been continuously updated — Vitalik Buterin's “Endgame” post (December 2021) sketched the destination, and his 2022 roadmap posts gave the plan its phase names. The Merge (2022) was the headline event, but it's only the first phase of a multi-year plan to scale Ethereum to tens of thousands of transactions per second without sacrificing its base layer's security or decentralization. Understanding the roadmap is understanding where the second-largest blockchain is actually going — and why its advocates believe rollups, not the base layer, are the future of crypto UX.

The roadmap in one sentence

Ethereum's plan, in its own framing: make the base layer a lean settlement and data-availability layer, push user transactions onto , and gradually fix the base layer's remaining problems (MEV, state bloat, client weight). Vitalik Buterin named the phases — the Merge, the Surge, the Scourge, the Verge, the Purge, the Splurge — in a series of 2022 roadmap posts that followed his late-2021 “Endgame” essay — and the Ethereum.org roadmap page is the canonical reference for the current status of each.

The Merge: proof-of-stake, finally

In September 2022, Ethereum switched from to . The immediate effects:

  • Energy use dropped ~99.95%. Validators replace miners; no more ASIC farms. The environmental argument against Ethereum largely evaporated.
  • Issuance fell. Staking rewards are far smaller than mining block rewards. Combined with EIP-1559's fee burn, Ethereum became net-deflationary during high-activity periods.
  • Staking and slashing. Validators lock up ETH and are penalized for misbehavior or downtime, with severe (“slashed”) penalties for acts that would compromise consensus — see the PoW vs PoS deep dive for the full trade-offs.

The Merge was a remarkable engineering feat — a live, $100B+ network changing its consensus mechanism without pausing. But it was always a prerequisite, not an end state. Scaling came next.

The timeline

Tap any event to expand its story.

The Surge: scaling via rollups and blobs

The central bet of the roadmap is that Ethereum's base layer will not scale to handle every user transaction. Instead, rollups execute transactions off-chain and post compressed data back to L1, inheriting L1's security. Two flavors dominate:

  • Optimistic rollups (Optimism, Arbitrum, Base): assume transactions are valid and allow a challenge window (with fraud proofs) before they finalize. Cheaper and simpler; the trade-off is a withdrawal delay.
  • ZK rollups (zkSync, Starknet, Linea, Scroll): post a cryptographic validity proof for every batch. Faster finality and no challenge window, but proving is more computationally expensive. See the Zero-Knowledge deep dive.

The 2024 Dencun upgrade (EIP-4844, “proto-danksharding”) introduced a new transaction type that carries “blobs” — large chunks of data priced separately from regular call data. Because rollups mostly need L1 for data availability (not computation), blob pricing dropped L2 fees by an order of magnitude overnight. Full danksharding (more, larger blobs) is the next step in the Surge.

Two further upgrades have shipped since. Pectra (May 7, 2025) roughly doubled blob capacity (a target of 6 blobs per block, max 9), raised a validator's maximum effective balance from 32 to 2,048 ETH, and introduced EIP-7702 — the first protocol-level account-abstraction upgrade, letting ordinary accounts temporarily adopt smart-contract code. Fusaka (December 3, 2025) brought PeerDAS, letting nodes verify blob data by sampling instead of downloading every blob — the first big step toward full danksharding.

The Scourge: MEV and censorship resistance

— the profit extractable by reordering, inserting, or censoring transactions — is one of Ethereum's most studied problems. After the Merge, a small number of builders and relays came to dominate block production, raising concerns about censorship (e.g., OFAC-compliant block production reached a majority share at one point). The Scourge's goal is to make MEV extraction less centralizing and censorship less feasible: mechanisms like inclusion lists (proposing validators can force certain transactions to be included) and encrypted mempools (transactions whose contents are hidden until inclusion) are the main candidates. For the background, see the MEV deep dive.

The Verge, the Purge, the Splurge

  • The Verge aims to make verifying the chain cheaper through stateless clients and Verkle trees, so that a node no longer needs to store the full state history. This directly helps decentralization: more people can run a full node.
  • The Purge is about pruning: removing old state data and protocol dead code so the chain stops growing forever. The goal is a stable, bounded state size.
  • The Splurge is the catch-all for everything else — smaller EIPs and long-tail improvements that don't fit a named phase. (Account abstraction, once slotted for here, has largely shipped: the ERC-4337 smart-account standard has been live since 2023, and Pectra's EIP-7702 brought it into the protocol in May 2025.)
Roadmaps slip — and that's normal

Ethereum's roadmap has always been optimistic on dates. The Merge was originally planned for 2017; it shipped in 2022. Full sharding was the original scaling plan; it was replaced by rollups + blobs when rollups matured faster. The lesson is not to trust specific dates, but to track which problems are structurally solved (proof-of-stake, blob data availability) and which are still open (full danksharding, stateless clients, MEV censorship resistance). The roadmap is a direction, not a schedule.

Why the base layer is being kept small

A recurring criticism is: why not just make the base layer bigger, like Solana? The Ethereum answer is that a large, high-throughput base layer is harder to run a full node for, which centralizes validation in data centers. The roadmap's bet — rollups for execution, a small base layer for settlement and data availability, plus Verge-era light verification — is specifically designed to preserve the property that ordinary users can verify the chain on consumer hardware. Whether that bet pays off is the central question for Ethereum's next decade.

Key takeaways

  • Ethereum's roadmap is a multi-year plan to turn the base layer into a lean settlement and data-availability layer, with user transactions pushed onto rollups. The phases (Merge, Surge, Scourge, Verge, Purge, Splurge) are named by Vitalik Buterin; Ethereum.org is the canonical status reference.
  • The Merge (Sep 2022) moved Ethereum to Proof of Stake, cutting energy use by ~99.95% and reducing issuance. It was a prerequisite, not the end state.
  • The Surge scales Ethereum via rollups posting data to L1. The 2024 Dencun upgrade (EIP-4844, proto-danksharding) added separately-priced blob space, dropping L2 fees by roughly an order of magnitude; Pectra (May 2025) roughly doubled blob capacity and shipped protocol-level account abstraction (EIP-7702), and Fusaka (December 2025) introduced PeerDAS.
  • The Scourge addresses MEV and censorship (post-Merge block production is concentrated; inclusion lists and encrypted mempools are the main candidates).
  • The Verge (stateless clients, Verkle trees), the Purge (state pruning), and the Splurge (long-tail EIPs) round out the plan. The roadmap's consistent goal is to preserve base-layer decentralization while scaling via L2s. For the competing scaling philosophy, see the PoW vs PoS deep dive.
Educational only, not financial or legal advice.