Open MLIR Meeting 11/2/2023: Proposal for a MPI Dialect

This Thursday, November 2nd (9am California Time, 16:00 UTC !! Some countries are in Winter time, we are still in the summer time this week !!), @AntonLydike will present a proposal for a MPI dialect.

MPI is at the heart of most modern massively parallel workloads, serving as the fundamental building blocks on which high-performance codes are built. With MLIR having proven itself well as a compiler for single-node workloads already, there is a lot interest in seeing MLIR scale up as a compiler for more massively parallel workloads. We are interested in bringing an MPI dialect into MLIR to serve as the foundation for a robust and efficient stack of dialects for distributed and even massively parallel workloads.

This presentation proposes a first patch that introduces a small subset of the MPI dialect. We want to discuss and review the design decisions made while creating this patch with the community and talk about the direction and scope of further development.

Zoom Meeting Link is unchanged, the presentation will be recorded and posted here and on our talks page on the website as usual.
You can also subscribe to this Google calendar to keep informed about incoming meetings.

Meeting ID: 851 5109 0498
Passcode: 828404


I tried to join but the zoom link got me into an empty meeting :cry:

The meeting was ~15 hours ago

1 Like

I was probably one hour early :face_with_open_eyes_and_hand_over_mouth:

Thanks all for joining, this was a fruitful discussion. Here are the slides and recording.

Sorry! I tried to call it out explicitly in the announcement:

Next time I’ll put it in bold!

Also: there is a public calendar that could help tracking this.

Thanks for the links to slides recording, I’ll have a look.

You did everything you could (even in bold!), it was entirely my inattention…

Thanks everyone again for the interest and all the good feedback! I’ve since written an RFC with much more information: [RFC] `MPI` Dialect