Our MLGO in LLVM panel was accepted at the 2022 LLVM Developers’ Meeting in San Jose, CA.
The goal is to bring together: compiler engineers working on ML-guided optimizations in LLVM, product engineers applying such optimizations in production, and researchers exploring the area.
A particular focus of this year’s discussions would be on deployment pain-points and lessons learned.
If you’re interested in participating, please send me a note[*] (name, affiliation, photo, attend remote/in-person) by Oct 9 (conference rules). Since the time is tight (total 40min), expect to have ~5 minutes to give an overview of your work - so that after all panelists do so, we can have sufficient time for a lively discussion.
(Side note - I’ll cancel the Oct meeting because no agenda and the Nov meeting because it’s right before this)
[*] If you already did, please just send me a note confirming and the remote/in person bit which we didn’t previously discuss, also by Oct 9