Ethereum Foundation

Fast Finality Workshop Cambridge

13–15 October 2026 · Jesus College

A working session on the Decoupled Consensus specification — closing its research chapter, and opening its review.

Goal

Decoupled consensus (DC) fixes known issues in today’s consensus, paves the way to a post-quantum-ready Ethereum, unlocks future networking and consensus upgrades, and speeds up finality out of the box.

This workshop’s goal is to align on the proposed protocol and specs, identify remaining problems, and kick off a new phase of the effort.

What begins here

Agenda

We tracked open problems on a board, by part of the protocol — available chain, stabilization gadget, finality gadget, healing, and networking. It filled up during red teaming, and the open sessions were scheduled from it to settle as much as we could. All times are local Cambridge time.

Watch recordings (opens in a new tab)Project site (opens in a new tab)PQ workshop (opens in a new tab)

  • DC design
  • Shipping DC & open sessions
  • Future of fast finality
  • Admin & misc

Monday

12

Arrivals

  1. 15:00Check-in opens
  2. 18:30Dinner at Murray Edwards College — organized, in plenum with the other workshops

Tuesday

13

The design, part one · consensus & networking

Watch day 1 (opens in a new tab)

  1. 08:30Doors open with coffee & pastries
  2. 09:00

    Intro

    Introductions · setting the stage · defining goals

    Leads Ben & Yann

    Watch recording: Intro (opens in a new tab)Slides: Intro (opens in a new tab)

  3. 09:30

    DC consensus: overview

    The consensus protocol, end to end

    Lead RobertoModerator Ben

    Watch recording: DC consensus: overview (opens in a new tab)Slides: DC consensus: overview (opens in a new tab)

  4. 11:00Coffee
  5. 11:30

    DC consensus: Goldilocks

    The consensus protocol, end to end

    Lead FrancescoModerator Ben

    Watch recording: DC consensus: Goldilocks (opens in a new tab)Slides: DC consensus: Goldilocks (opens in a new tab)

  6. 12:30Lunch
  7. 14:00
  8. 15:00
  9. 16:00Coffee
  10. 16:30

    Future of fast finality: PQ networking

    Aggregator design

    Leads Kev & KamilModerator Ben

    Watch recording: Future of fast finality: PQ networking (opens in a new tab)

  11. 17:00

    Open session I

    Goldilocks vs. the full stabilization gadget · the locking period

    Watch recording: Open session I (opens in a new tab)

  12. Dinner — individual

Wednesday

14

The design, part two · future of fast finality

Watch day 2 (opens in a new tab)

  1. 08:30Doors open with coffee & pastries
  2. 09:00

    Recap

    Decisions from day one · which open problems to tackle today

    Watch recording: Recap (opens in a new tab)

  3. 09:15

    Open sessions II–III (consensus)

    Goldilocks vs. the full stabilization gadget · VRF or non-VRF available chain · healing · honest leaking under low participation

    Watch recording: Open sessions II–III (consensus) (opens in a new tab)Slides: Open sessions II–III (consensus) (opens in a new tab)

  4. 11:00Coffee
  5. 11:30

    Incentives

    Rewards, penalties, participation · inactivity leak

    Lead AndersModerator Ben

    Watch recording: Incentives (opens in a new tab)Slides: Incentives (opens in a new tab)

  6. 12:30Lunch
  7. 14:00

    Specs, prototyping & the transition

    Specs, tests, devnets · from today’s consensus to decoupled consensus

    Leads Mikhail, Justin & Thomas CModerator Ben

    Watch recording: Specs, prototyping & the transition (opens in a new tab)Slides: Specs, prototyping & the transition (opens in a new tab)

  8. 15:00

    Future of fast finality: PQ networking

    PQ strategy · heartbeat · leanVM · aggregator design · optimization

    Leads Kev & KamilModerator Ben

    Watch recording: Future of fast finality: PQ networking (opens in a new tab)Slides: Future of fast finality: PQ networking (opens in a new tab)

  9. 15:45

    Tens-of-seconds finality with PQ validators

    Bounds on fast finality · PQ validator set management · validator priority · consolidation

    Lead BarnabéModerator Ben

    Watch recording: Tens-of-seconds finality with PQ validators (opens in a new tab)Slides: Tens-of-seconds finality with PQ validators (opens in a new tab)

  10. 16:30Coffee
  11. 17:00

    Lean clients

    Experience so far · future direction

    Leads Mihir, Mega, Pablo & Locki

    Watch recording: Lean clients (opens in a new tab)

  12. 17:45

    Open sessions IV–V (networking)

    Aggregation bitfields · advancing heights during a round · 4-slot finality in I* · staggered publishing vs. multiple committees

    Watch recording: Open sessions IV–V (networking) (opens in a new tab)Slides: Open sessions IV–V (networking) (opens in a new tab)

  13. 19:00Drinks reception at Trinity Hall — Terrace Room
  14. 19:30Dinner — Graham Storey Room, until 23:00

Thursday

15

The road to review

Watch day 3 (opens in a new tab)

  1. 08:30Doors open with coffee & pastries
  2. 09:00
  3. 12:30Lunch
  4. 14:30

    Open session VI (networking)

    Gossip priorities in Prysm on constrained uplinks

    Lead Csaba

    Watch recording: Open session VI (networking) (opens in a new tab)Slides: Open session VI (networking) (opens in a new tab)

  5. 15:00

    ACDC

  6. 16:30Coffee
  7. 17:00

    Future of fast finality: 1RF

    One-round finality · safety gadget · optimistic path · fast/slow path · lean Ethereum · the extremely lean chain

    Leads Francesco & YannModerator Ben

    Watch recording: Future of fast finality: 1RF (opens in a new tab)Slides: Future of fast finality: 1RF (opens in a new tab)

  8. Security review

    What do we need to stabilize DC? Academic review · core dev review · independent formal verification · incentive analysis · DoS vectors · network assumptions · aggregator assumptions

    Lead YannModerator Ben

  9. Synthesis

    FV workshop & open sessions: what’s settled, what’s still open

    Moderator Ben

  10. Conclusions

    Working group · commitments · next steps

    Lead Ben

    Flipchart summary of open issues (opens in a new tab)

  11. Dinner — individual

Friday

16

Farewell

  1. 09:00Check-out
  2. 09:30Optional social event, decided spontaneously
  3. 12:30Departures

Background

The proposal is public: “Unblocking faster finality with decoupled consensus” (ethresear.ch, March 2026).

Resources

EIP-8453: Decoupled Consensus (draft pull request; discussion on Ethereum Magicians)

Decoupled consensus links (specs, code and background reading, collected by Ben on HackMD)

Practicalities

When
Three full days, 13–15 October 2026 (Tuesday to Thursday)
Where
Sibilla Room, Jesus College, Cambridge