Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Execution Layer Meeting 194 #1124

Closed
timbeiko opened this issue Aug 1, 2024 · 10 comments
Closed

Execution Layer Meeting 194 #1124

timbeiko opened this issue Aug 1, 2024 · 10 comments

Comments

@timbeiko
Copy link
Collaborator

timbeiko commented Aug 1, 2024

Meeting Info

Agenda

@potuz
Copy link

potuz commented Aug 3, 2024

I'd like to spend like 5 minutes pitching EIP-7732. It's part of the formal process depicted in EIP-1 and ACDE is the right place and time for it: most CL devs and researchers are already aware and have opinions on it, EL devs weren't mostly approached as the EIP is purely a consensus change, but it has strong implications on how and when the execution payload is validated. The timing is right because the spec is in a somewhat stable form and implementation is picking up pace at the moment, it's a good time to raise awareness and catch fatal flaws before they take back months of implementation

@Giulio2002
Copy link

I would like to pitch ethereum/execution-apis#570. It allows clients to remove totalDifficulty from the jsonrpc, thus allowing for refactoring of pre-merge code-debts.

@MariusVanDerWijden
Copy link
Member

I would like to discuss eth/69: Dropping pre-merge fields from the network messages: https://eips.ethereum.org/EIPS/eip-7642

It's a small EIP that will greatly reduce the bandwith required during sync (~530 GB)

@MariusVanDerWijden
Copy link
Member

MariusVanDerWijden commented Aug 15, 2024

We also recently figured out (thanks to @jwasinger) that our library for bls12-381 (gnark) uses all available cores during the bls multiexponentiations. Thus the benchmarks we did for gas costs are skewed.

I get 18-30MGas for G1 compared to 27-60MGas:
For G2 I get 40-50MGas compared to 78-90MGas.
For comparison Ecrecover does 38-44 MGas on my machine.

Seems like we might need to reprice the G1 Multiexp for Pectra.

It would be nice if other clients could also run some bls benchmarks to see where their current costs are at

@tersec
Copy link

tersec commented Aug 15, 2024

I'd like to discuss rebasing PeerDAS on Pectra. This has implications with regard to testing the EL-centric EIPs 7702 and EOF.

@holgerd77
Copy link

I would like to discuss eth/69: Dropping pre-merge fields from the network messages: https://eips.ethereum.org/EIPS/eip-7642

Just read through the EIP, support from EthereumJS!

We also recently figured out (thanks to @jwasinger) that our library for bls12-381 (gnark) uses all available cores during the bls multiexponentiations. Thus the benchmarks we did for gas costs are skewed.

That's a great find! 🤩

Maybe we should make it a standard procedure to always have such benchmark calculations done by two different client implementations/teams min for future EIPs to minimize the risk of such side effects remain undiscovered? 🤔

@barnabasbusa
Copy link
Member

Quick reminder, we going to have weekly Pectra Interop testing call starting from next week Monday starting at 14:00 UTC.

@poojaranjan
Copy link
Contributor

Recording - https://www.youtube.com/watch?v=tbxgYq8KmmM

@akashkshirsagar31
Copy link

Podcast (audio only) - https://open.spotify.com/episode/4QUOJ7u9KuwSRbJJyrWPHZ?si=QafZbaXBQbCwJ3HcaCxqkg

Aleneth added a commit to Aleneth/pm that referenced this issue Aug 20, 2024
@timbeiko
Copy link
Collaborator Author

Closed in favor of #1142

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

9 participants