Ethereum Foundation
Fast Finality WorkshopCambridge
13–15 October 2026 · Jesus College
Agenda
Password
weeping-amuze-bejan-shoelace-burrknotScan to open the agenda
Open problems
The board
Add anything you’d like to bring up during the open sessions, or things you’d like to challenge.
notes.ethereum.org/yNWbJIsoROGSdT3Nq04Tpw · also linked from the agenda page
Scan to open the board
Introductions
- Your name, and where you work
- What you are working on
- What you want to leave Cambridge with
Expectations
What we’re doing is very hard
Tuesday · 09:00
Very confused
Thursday · 17:30
Aligned
These are (almost) unique to us.
…but they also require:
Context
How we got here, and what’s next
2022 – 2025
Years of research
- Goldfish
- RLMD‑GHOST
- TOB‑SVD
- 3SF
- Majorum, …
Oct 2025Cambridge
Not BFT‑only
+ stabilization gadget
Nov 2025Devconnect ARG
First designs validated
Jan 2026Berlin
Strawmap: decouple DC & 1RF
Mar 2026ethresear.ch
DC goes public
Oct 2026Cambridge
Converge
You are here
3–6 Nov 2026Mumbai
Devcon
In three weeks, in front of core devs
Could startsoon
I* headliner selection
So, this week
- Don’t give core devs something unready to look at.
- Add Lindy-ness to the right thing.
Programming
Three days, three objectives
- DC design Objective 1
- Shipping DC & open sessions Objective 2
- Future of fast finality Objective 3
- Formal verification
- Admin & misc
Tuesday
The design, part one · consensus & networking
- 09:00Intro
- 09:30DC consensus: overviewRoberto
- 11:30DC consensus: GoldilocksFrancesco
- 14:00Goldilocks, continuedFrancesco
- 15:00DC networking: design & benchmarkingSukun
- 16:30PQ networkingKev & Kamil
- 17:00Open session I
Wednesday
The design, part two · future of fast finality
- 09:00Recap
- 09:15Open sessions II–III · consensus
- 11:30IncentivesAnders
- 14:00Specs & prototypingMikhail, Justin & Thomas C
- 14:30The transitionMikhail & Justin
- 14:45PQ networkingKev & Kamil
- 15:15Lean clientsMihir, Mega, Pablo & Locki
- 15:45Tens-of-seconds PQ finalityBarnabé
- 17:00Open sessions IV–V · networking
Thursday
The road to review
- 09:00Formal verification · with the FV workshopRoberto, Francesco & Alex
- 14:00One-round finalityFrancesco & Yann
- 14:30Open session VI
- 15:00ACDC
- 17:00Security reviewYann
- 17:15Synthesis
- 17:30Conclusions · working group, commitments, next steps
Reminder
Why we need decoupled consensus
Post-quantum readiness
A small committee signs with post-quantum signatures as they are. The heartbeat goes PQ first; finality follows.
Fixing Gasper
A dynamically available, reorg-resilient chain in place of LMD-GHOST, cleanly separated from finality.
Fast finality
Finality runs in parallel, off the critical path, so slots and finality no longer trade off.
Starting point
Decoupled consensus, as it stands
Let’s see how much stays the same — and how much we can improve.
I’ll try to track the protocol here as it evolves.
One transaction, from inclusion to finality
≈ 3m 48sexpected · today ≈ 16 min
Expected values · finality.watch “DC today” model · ~1.1M validators · 12-second slots
Goals
