- mpi-forum - mailman3.cels.anl.gov

Re: [Mpi-forum] Non-Voting Meeting - September 10th - Dynamic Process Sets
by Wes Bland 08 Sep '25

08 Sep '25

Reminder that we have a non-voting meeting this Wednesday from the Sessions WG to discuss dynamic process sets. Connection information in the usual place on the wiki: https://urldefense.us/v3/__https://github.com/mpi-forum/mpi-standard/wiki/V… If that link doesn’t work for you, it’s because you’re not signed into your GitHub account and/or you aren’t a member of the GitHub group. Please let me know if you need access. Thanks, Wes > On Aug 26, 2025, at 10:24 AM, Wes Bland <WORK(a)WESBLAND.COM> wrote: > > Hi all, > > Now that we’re done with MPI 5.0, it’s time to start working on larger changes that were pushed out a bit while we finished the ABI. > > To kick that off, Howard Pritchard, Dominik Huber, and the Sessions WG have requested some time before the upcoming Charlotte meeting to discuss dynamic process sets. > > We’ll meet on Wednesday, September 10th at 10am US Central time <https://urldefense.us/v3/__https://www.worldtimebuddy.com/?qm=1&lid=6,5,264… > (click link for your local time). We reserve two hours for these meetings in case we need the extended time. > > As always, you can subscribe to the iCal file to get this on your calendar app of choice: > > https://urldefense.us/v3/__https://calendar.google.com/calendar/ical/g5ms8r… > > Thanks, > Wes

1 0

Non-Voting Meeting - September 10th - Dynamic Process Sets
by Wes Bland 26 Aug '25

26 Aug '25

Hi all, Now that we’re done with MPI 5.0, it’s time to start working on larger changes that were pushed out a bit while we finished the ABI. To kick that off, Howard Pritchard, Dominik Huber, and the Sessions WG have requested some time before the upcoming Charlotte meeting to discuss dynamic process sets. We’ll meet on Wednesday, September 10th at 10am US Central time <https://urldefense.us/v3/__https://www.worldtimebuddy.com/?qm=1&lid=6,5,264… > (click link for your local time). We reserve two hours for these meetings in case we need the extended time. As always, you can subscribe to the iCal file to get this on your calendar app of choice: https://urldefense.us/v3/__https://calendar.google.com/calendar/ical/g5ms8r… Thanks, Wes

1 0

Re: [Mpi-forum] Request for input on Libfabric experiences
by Derek Schafer 22 Aug '25

22 Aug '25

Hello again! We'd like to thank everyone who has filled our survey so far. We're looking to close this survey in about a week, so we would like to gently encourage those with even a slight interest in Libfabric to take a few minutes and fill out our survey. Survey Link<https://urldefense.us/v3/__https://docs.google.com/forms/d/e/1FAIpQLSc-xgbg… > Thank you, Derek From: mpi-forum <mpi-forum-bounces(a)lists.mpi-forum.org> On Behalf Of Derek Schafer via mpi-forum Sent: Thursday, August 14, 2025 10:30 AM To: mpi-forum(a)lists.mpi-forum.org Subject: [Mpi-forum] Request for input on Libfabric experiences [EXTERNAL] Hello MPI Forum members! Our group, the Center for Understandable, Performant Exascale Communication Systems, has been focused on providing optimized, performance-transparent communication systems for the last five years. Recently, we have observed ZjQcmQRYFpfptBannerStart This Message Is From an External Sender This message came from outside your organization. ZjQcmQRYFpfptBannerEnd Hello MPI Forum members! Our group, the Center for Understandable, Performant Exascale Communication Systems, has been focused on providing optimized, performance-transparent communication systems for the last five years. Recently, we have observed an increased interest in Libfabric, and would like to offer one or potentially multiple tutorials on this topic. If such a tutorial might be of interest to you or your institution, we kindly ask that you fill out this google forms survey<https://urldefense.us/v3/__https://docs.google.com/forms/d/e/1FAIpQLSc-xgbg…>. We are hoping to collect input from a wide range of experiences with Libfabric so we can best address the needs and interests of the community with these tutorials. Also, if you have other colleagues who may be interested in such a tutorial, or would have other insights, we would appreciate if you could forward this email to them as well. If anyone has any issues filling out the survey, please feel free to reach out to me directly. Kind regards, Derek Schafer on behalf of Drs. Patrick Bridges, Amanda Bienz, Puri Bangalore, and Anthony Skjellum Center for Understandable, Performant Exascale Communication Systems --- Derek Schafer (dschafer1(a)unm.edu<mailto:[email protected]>) Research Scientist Center for Advanced Research Computing University of New Mexico

1 0

Request for input on Libfabric experiences
by Derek Schafer 14 Aug '25

14 Aug '25

Hello MPI Forum members! Our group, the Center for Understandable, Performant Exascale Communication Systems, has been focused on providing optimized, performance-transparent communication systems for the last five years. Recently, we have observed an increased interest in Libfabric, and would like to offer one or potentially multiple tutorials on this topic. If such a tutorial might be of interest to you or your institution, we kindly ask that you fill out this google forms survey<https://urldefense.us/v3/__https://docs.google.com/forms/d/e/1FAIpQLSc-xgbg… >. We are hoping to collect input from a wide range of experiences with Libfabric so we can best address the needs and interests of the community with these tutorials. Also, if you have other colleagues who may be interested in such a tutorial, or would have other insights, we would appreciate if you could forward this email to them as well. If anyone has any issues filling out the survey, please feel free to reach out to me directly. Kind regards, Derek Schafer on behalf of Drs. Patrick Bridges, Amanda Bienz, Puri Bangalore, and Anthony Skjellum Center for Understandable, Performant Exascale Communication Systems --- Derek Schafer (dschafer1(a)unm.edu<mailto:[email protected]>) Research Scientist Center for Advanced Research Computing University of New Mexico

1 0

Call for agenda items for the MPI Forum in September, Deadline September 15th
by Martin Schulz 12 Aug '25

12 Aug '25

Hi all, The next MPI forum meeting is getting closer (see also Joseph’s call for registration for forum and EuroMPI the other day). The forum is scheduled to take place before the EuroMPI conference in Charlotte, this means from September 29 to September 30 or October 1, depending on demand. We will do this, as usual, in a hybrid fashion, but – of course – we hope that many/most of you can make it in person. With MPI 5.0 complete, which took most of our attention, we need to shift to new topics (or old, still pending topics) towards MPI 6.0. In this part, I do expect less plenary topics in terms of tickets and ballots, but if you do have something, please announce it to the list by September 15th at the latest (earlier is, of course, appreciated), and we will add it to the agenda. For the remaining time, I was hoping that we can re-activate our tradition of having longer WG breakouts at the forum meetings, as well as more forward looking discussions on MPI-next in topical plenary slots. If you have particular topics that you would like to bring up, or if your WG would like a slot for a breakout, please let us know. The sooner, the better, so we can assemble an agenda. Thanks! Martin -- Prof. Dr. Martin Schulz, Chair of Computer Architecture and Parallel Systems Department of Informatics, TU-Munich, Boltzmannstraße 3, D-85748 Garching Member of the Board of Directors at the Leibniz Supercomputing Centre (LRZ) Email: schulzm(a)in.tum.de You never fail until you stop trying. (Albert Einstein)

1 0

EuroMPI & MPI Forum registration
by Joseph Schuchart 10 Aug '25

10 Aug '25

1 0

Re: [Mpi-forum] Proposal process for a side document for C++ Language Interface for MPI-5.x
by Wes Bland 01 Jul '25

01 Jul '25

Hi Tony, I’ll refer to the procedures doc here: https://urldefense.us/v3/__https://www.mpi-forum.org/docs/other/procedures-… (*pushes glasses up nose* - Section 2.2.5). For side documents, you’ll follow the same procedure as a regular proposal (have a plenary/reading first, then two votes). The only difference is that updates to side docs incrementing the minor version number only have to pass a single vote (like an errata). Organizationally, we have repositories for each of the side documents in the main MPI Forum GitHub organization so it’s easier for folks to find them. You can find an example with the newer Memory Allocation Kinds side document here: https://urldefense.us/v3/__https://github.com/mpi-forum/mem-alloc__;!!G_uCf… . I’d be happy to make a repo for you if you’re interested. I'll put you on the list for the North Carolina meeting. I’m also happy to help schedule a non-voting meeting if you want to get some early feedback during one of our Wednesday time slots. Thanks, Wes > On Jun 26, 2025, at 10:10 AM, Skjellum, Anthony via mpi-forum <mpi-forum(a)lists.mpi-forum.org> wrote: > > This Message Is From an External Sender > This message came from outside your organization. > Dear Wesley and Martin, how is a side document proposed to the forum? > > We wish to prepare a side document. > > Also, here is our recent EuroMPI paper describing the thinking and learning of the subcommittee on C++; we would also like to have a chance to share this at the SC25 BOF. > > If we can have plenary time at the next meeting, that would be good. North Carolina, right? > > Here is the EuroMPI paper from Arxiv: > > https://urldefense.us/v3/__https://www.arxiv.org/abs/2506.14610__;!!G_uCfsc… <https://urldefense.us/v3/__https://www.arxiv.org/abs/2506.14610__;!!G_uCfsc…> > > > Thanks, > Tony > > > Anthony Skjellum, PhD > Professor of Computer Science > Director, Advanced Scalable Computing, > Extreme Networks & Data (ASCEND) Center > Tennessee Technological University > email: askjellum(a)tntech.edu <mailto:[email protected]> > cell: +1-205-807-4968 > > > _______________________________________________ > mpi-forum mailing list > mpi-forum(a)lists.mpi-forum.org <mailto:[email protected]> > https://urldefense.us/v3/__https://lists.mpi-forum.org/mailman/listinfo/mpi…

1 0

Proposal process for a side document for C++ Language Interface for MPI-5.x
by Skjellum, Anthony 26 Jun '25

26 Jun '25

Dear Wesley and Martin, how is a side document proposed to the forum? We wish to prepare a side document. Also, here is our recent EuroMPI paper describing the thinking and learning of the subcommittee on C++; we would also like to have a chance to share this at the SC25 BOF. If we can have plenary time at the next meeting, that would be good. North Carolina, right? Here is the EuroMPI paper from Arxiv: https://urldefense.us/v3/__https://www.arxiv.org/abs/2506.14610__;!!G_uCfsc… Thanks, Tony Anthony Skjellum, PhD Professor of Computer Science Director, Advanced Scalable Computing, Extreme Networks & Data (ASCEND) Center Tennessee Technological University email: askjellum(a)tntech.edu cell: +1-205-807-4968

1 0

MPI 5.0 Voted In
by Martin Schulz 05 Jun '25

05 Jun '25

Hi all, MPI 5.0 has been voted in and the final, ratified document has been published on our website (thanks to Bill and Wes for finishing this up so quickly!): https://urldefense.us/v3/__https://www.mpi-forum.org/docs/mpi-5.0/mpi50-rep… Congratulations everyone and also a big think you to each and everyone of you in the MPI Forum who contributed to the document. Special thanks to Jeff Hammond and the ABI WG, who pushed the big change for this version, the ABI! Please all help promote the new standard – I just put this on linked-in and will try to also find other venues, but feel free to also post, repost or disseminate in any other way. Let’s get the word out. Also, if you happen to be in Hamburg next week for ISC, please drop me a note, so we can also coordinate the activities there. In other news, all four officers have been re-elected – thank you for your continued trust and support. Now, on to MPI 5.1 or MPI 6.0, depending on what we decide going forward – I am looking forward to this next chapter in MPI and our continued work in the MPI Forum! Martin -- Prof. Dr. Martin Schulz, Chair of Computer Architecture and Parallel Systems Department of Informatics, TU-Munich, Boltzmannstraße 3, D-85748 Garching Member of the Board of Directors at the Leibniz Supercomputing Centre (LRZ) Email: schulzm(a)in.tum.de You never fail until you stop trying. (Albert Einstein)

1 0

Re: [Mpi-forum] MPI_Psend/recv_init_c Resolution
by Wes Bland 05 Jun '25

05 Jun '25

Hi all, I discovered this morning that this PR was never merged into the repository, which means it wasn’t in the RC doc that we’ve been reviewing. There was no objection at the time and this was never going to be voted on, so I’m not proposing that we add any new votes now, but I wanted to call people’s attention to the fact that the mistake hadn’t been corrected yet, but I’m going to merge it now. So the doc that we’d publish as MPI 5.0 would include everything that was already there, plus reverting this mistake that I added at the March meeting. Sorry again for the confusion. Thanks, Wes > On Mar 21, 2025, at 11:30 AM, Wes Bland via mpi-forum <mpi-forum(a)lists.mpi-forum.org> wrote: > > This Message Is From an External Sender > This message came from outside your organization. > Hi folks, > > During the meeting in Stuttgart, we discovered some weirdness where we thought MPI_Psend_init_c and MPI_Precv_init_c should have been present in MPI 4.1, but weren’t. Failing to find any reason why that was true, we assumed it was a problem with a Git merge and I replayed those changes in PRs 1108 <https://urldefense.us/v3/__https://github.com/mpi-forum/mpi-standard/pull/1…> and 1110 <https://urldefense.us/v3/__https://github.com/mpi-forum/mpi-standard/pull/1…>. > > Yesterday, Ken found the place where we had discussed these specific topics in October, 2023 and decided not to put them in MPI 4.1 here <https://urldefense.us/v3/__https://github.com/mpi-forum/mpi-issues/issues/7…>. At that time, we decided to push them out of 4.1 and address them later (which we never circled back to do) and removed the changes here <https://urldefense.us/v3/__https://github.com/mpi-forum/mpi-standard/pull/9…>. > > Now that we know what happened there, I’m going to revert the PRs to put these changes back in and put us back to the situation we were in before the RCM and put this back in the Forum’s queue to discuss at some point post-MPI 5.0. You can find the PR for that here <https://urldefense.us/v3/__https://github.com/mpi-forum/mpi-standard/pull/1…>. > > I’ll leave that up for a few days before proposing it gets merged just so everyone has a chance to take a look. > > Sorry about all the mess. Hopefully we’re all back to the point that we previously agreed on. > > Thanks, > Wes > _______________________________________________ > mpi-forum mailing list > mpi-forum(a)lists.mpi-forum.org <mailto:[email protected]> > https://urldefense.us/v3/__https://lists.mpi-forum.org/mailman/listinfo/mpi…

1 0