EPF Week 8 Updates

Topic: PEPC-Boost Research

This week we had 90min call with our PEPC-Boost core team that consists of Barnabé, Diego, Bharath, and me. Now that the PEPC-Boost design is more clear, we figured there are four main research questions that we want to answer. This is currently my main focus.

We are going to have biweekly 90min sync calls on our progress.

PEPC-Boost open questions and my focus:
1. Information available to ROB builder: “What is the data available to the ROB builder when they are building? They know the pre-state of the TOB, which is the post-state of the block in the previous slot, but they don’t know the contents of the TOB. What are the assumptions they can/should make about it?”
2. TOB+ROB assembly: (related to 1.) “most conservative: ROB isn't allowed to touch the same addresses. Least conservative: ROB is just appended as is, without care for reversions. But is there something in between that has low latency and high efficiency?”
3. Payments (related to 1. and 2.)
4. Connecting to PEPC-Boost: What strategies for the proposer?

1. Players: 1 proposer + m builders
2. Actions: There are n relays, and each player can connect to a set of them, i.e., each of the
   $m+1$ players has an action set equal to
   $2^n$ (set of relays they connect to)
3. Utilities: These are unknown. So we need more research on questions 1-3. to think through that game properly.

Related materials

• Why enshrine Proposer-Builder Separation? A viable path to ePBS. [Mike Neuder, Justin Drake]
• PEPC FAQ [Barnabé Monnot]
• Ethereum PBS Roadmap [Mike Neuder]
• Optimistic relays and where to find them
• Equivocation attacks on ePBS
• Why unbundling the auction makes sense
• Some stats around MEV extraction from CEX/DEX arbs
• Cosmos design for TOB auctions, potentially relevant/good to see how they do it elsewhere/project to track