Home
Keep track of advancements in every frontier

TheTechno Optimist.

Ethereum protocol upgradesEthereum Foundation—Lean EthereumEthereum Foundation—Ethereum protocol upgradesEthereum Foundation—Lean EthereumEthereum Foundation—

Ethereum Foundation

Ethereum research and upgrades

CHprivateest. 2014ethereum.foundation (opens ethereum.foundation)Checked 25 Sep

The brief

Analyst view
  • The Merge (15 Sept 2022) switched Ethereum to proof of stake, cutting energy use by ~99.9%.
  • Fusaka (Dec 2025) lets nodes check rollup data by sampling pieces instead of downloading it all.
  • Glamsterdam, the next upgrade, is on test networks, with mainnet targeted for late 2026.

Technical approach

As reported
Consensus

Staked validators vote every 12 seconds

Validators lock up ETH and vote on each block. Once two-thirds of the stake agrees, blocks become final, usually after ~13 minutes.

Scaling

Rollup-first roadmap

Most activity moves to rollups that post compressed data to Ethereum. Ethereum focuses on storing that data and settling results.

Data availability

Blobs and data sampling

Blobs are cheap, temporary data slots for rollups (EIP-4844). With PeerDAS, each node keeps only a few pieces yet can tell if data is missing.

Clients

Many independent clients

About 10 independent software clients, such as Geth and Lighthouse, follow the same rules. A bug in one doesn't halt the chain.

Programs

Ethereum Foundation, two programs

Updated 3 Sep 2026
Ethereum protocol upgradesThe planned upgrades ('hard forks') that change Ethereum's rules. Recent ones moved to proof of stake and made rollup data cheaper.ScalingNextGlamsterdam: built-in block auctions (ePBS) and access lists · Dec 2026+6 moUpdated 3 Sep 2026View program →Lean EthereumA long-term plan to rebuild Ethereum's core with post-quantum signatures, faster finality and blocks checked by ZK proofs.ResearchNextPrototype clients and test networksUpdated 13 Aug 2026View program →
Research · Proto · Pilot · Deploy · Scaling

Research & demos

4 items

Open technical problems

Open · Active · Partial
  • Block building is concentratedActive research

    A few specialist firms build most blocks to capture trading profits (MEV). Moving the auction into the protocol (ePBS) won't fix that alone.

  • Slow finalityActive research

    ~13 minutes is long for exchanges and bridges. Finality in one slot means handling votes from a huge validator set every slot.

  • Quantum-resistant signaturesOpen

    Large quantum computers could break today's signatures. Swapping in post-quantum ones without bloating blocks is still research.

Upcoming catalysts

Company-wide
  • 27 Oct2026Glamsterdam on the Hoodi testnet (tentative date from core-dev calls)Ethereum protocol upgradesin 17 d
  • 2027Production-grade leanVM targeted, per Justin Drake; layer deployments planned from 2028Lean Ethereumby ~15 mo

Peers

Blockchains (L1s) · 1

Company & funding

HQ
CH
Founded
2014

Partners & customers

3 partners
  • Partner
    Coinbase (Base)

    Base is the largest L2 by activity, settling to Ethereum

  • Partner
    Eigen Labs

    EigenLayer restaking reuses staked ETH to secure other services

  • Partner
    StarkWare

    Starknet validity rollup settles to Ethereum

People

Key people
Primer