- mpiwg-coll - mailman3.cels.anl.gov

PR for scan/exscan clarification
by Victor Eijkhout 10 Jun '26

10 Jun '26

Hello collective, Can I have an opinion on github.com<https://urldefense.us/v3/__https://github.com/mpi-forum/mpi-standard/pull/1… > [X]<https://urldefense.us/v3/__https://github.com/mpi-forum/mpi-standard/pull/1… > which brings the syntax diagram of a few scan routines in line with other collectives, and improves on the accompanying text? Victor.

1 0

Review of figure changes in chapter 6
by Benson Muite 22 May '26

22 May '26

1 0

Re: [mpiwg-coll] New meeting of MPI Forum Collective WG
by Julien.JAEGER＠CEA.FR 20 May '26

20 May '26

We are meeting today in 30 minutes. Julien Jaeger. ________________________________ De : Skjellum, Anthony <askjellum(a)tntech.edu> Envoyé : mercredi 22 avril 2026 15:49:40 À : mpiwg-coll(a)lists.mpi-forum.org; mpiwg-persistence(a)lists.mpi-forum.org Cc : Raffenetti, Ken; JAEGER Julien 606512; Dosanjh, Matthew Gurbachan Fugelso; Bengisu Elis Objet : Re: New meeting of MPI Forum Collective WG We are meeting today. 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 ________________________________ From: Skjellum, Anthony Sent: Friday, February 6, 2026 10:26 AM To: mpiwg-coll(a)lists.mpi-forum.org <mpiwg-coll(a)lists.mpi-forum.org>; mpiwg-persistence(a)lists.mpi-forum.org <mpiwg-persistence(a)lists.mpi-forum.org> Cc: Raffenetti, Ken <raffenet(a)anl.gov>; Julien.JAEGER(a)CEA.FR <Julien.JAEGER(a)CEA.FR>; Dosanjh, Matthew Gurbachan Fugelso <mdosanj(a)sandia.gov>; Bengisu Elis <bengisu.elis(a)tum.de> Subject: New meeting of MPI Forum Collective WG When: Wednesday, April 22, 2026 9:00 AM-10:00 AM. Where: Via Zoom All: Julien Jaeger and I are resuming. Agenda forthcoming. Will be every other week, to dovetail with RMA WG. Tony Skjellum (TNTECH) is inviting you to a scheduled Zoom meeting. Topic: MPI WG Collective (Persistent too) Time: This is a recurring meeting Meet anytime Join Zoom Meeting https://urldefense.us/v3/__https://tntech.zoom.us/j/85895299644?pwd=UVeoehU… Meeting ID: 858 9529 9644 Passcode: mpi4eva --- One tap mobile +13017158592,,85895299644#,,,,*2477715# US (Washington DC) +13052241968,,85895299644#,,,,*2477715# US Join instructions https://urldefense.us/v3/__https://tntech.zoom.us/meetings/85895299644/invi…

1 0

Re: [mpiwg-coll] New meeting of MPI Forum Collective WG
by Julien.JAEGER＠CEA.FR 25 Mar '26

25 Mar '26

Dear all, As Tony and I are not available, today's meeting is canceled. We will meet in two weeks. If you have a specific subject you want to bring up, discuss or present in a coming meeting, do not hesitate to send a email to Tony and I so we can update the running agenda accordingly. Best regards, Julien Jaeger. ________________________________ De : mpiwg-persistence <mpiwg-persistence-bounces(a)lists.mpi-forum.org> de la part de Skjellum, Anthony via mpiwg-persistence <mpiwg-persistence(a)lists.mpi-forum.org> Envoyé : mercredi 11 février 2026 15:47:20 À : mpiwg-coll(a)lists.mpi-forum.org; mpiwg-persistence(a)lists.mpi-forum.org Objet : Re: [mpiwg-persistence] New meeting of MPI Forum Collective WG Agenda for today (this will be on-going, and we will add to/resolve over time) AGENDA - FEB 11, 2026 Reconstitution - subsume partitioned, persistent, point-to-point and collective discussions Items for good of the order form participants not ZjQcmQRYFpfptBannerStart This Message Is From an External Sender This message came from outside your organization. ZjQcmQRYFpfptBannerEnd Agenda for today (this will be on-going, and we will add to/resolve over time) AGENDA - FEB 11, 2026 1. Reconstitution - subsume partitioned, persistent, point-to-point and collective discussions 2. Items for good of the order form participants not covered below 3. Align with hybrid/streams work in pt2pt scope BIG PICTURE AGENDA FOR GROUP, NOT EVERYTHING FOR FIRST MEETING. 1. Discuss partitioned collective status and plans * Prior papers * Issues with synchronizing collectives * Prototypes * New ideas with buffer management * Path to standardization 2. Corollary issues with persistent collectives * Non-blocking init/free - not allowed 3. Related issue - handoff of control of user state * Handling the registration and registration of index vectors and data pointers * C++ way * C (and Fortran) way with registration callbacks at free * Removing ambiguity in the “zero time use case” and whenever a free is allowed to lazy in MPI that holds user state (so it can be released before MPI_Finalize). 1. Managing unexpected state in partitioned (and maybe persistent) operations * Elimination of PBUF_PREPARE in all partitioned considerations * Considering to “generalize” point-to-point partitioned to enable a) persistence better; b) buffer pools 1. Streaming collectives 2. Other things we “owe” short term * MPI_Parrived_any ticket — very useful with current partitioned point-to-point 3. Document enhancements * Revamping the collective chapter front-to-back (ask Puri if he had some input on prior tries at this?) * Making uniform the presentation of blocking, nonblocking, persistent, and future (partitioned) collective * Front matter needs a look for terms, definitions… --- Running minutes will be captured and shared. Recording whenever possible :-) 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 ________________________________ From: Skjellum, Anthony Sent: Friday, February 6, 2026 10:26 AM To: mpiwg-coll(a)lists.mpi-forum.org <mpiwg-coll(a)lists.mpi-forum.org>; mpiwg-persistence(a)lists.mpi-forum.org <mpiwg-persistence(a)lists.mpi-forum.org> Cc: Raffenetti, Ken <raffenet(a)anl.gov>; Julien.JAEGER(a)CEA.FR <Julien.JAEGER(a)CEA.FR> Subject: New meeting of MPI Forum Collective WG When: Wednesday, February 11, 2026 9:00 AM-10:00 AM. Where: Via Zoom All: Julien Jaeger and I are resuming. Agenda forthcoming. Will be every other week, to dovetail with RMA WG. Tony Skjellum (TNTECH) is inviting you to a scheduled Zoom meeting. Topic: MPI WG Collective (Persistent too) Time: This is a recurring meeting Meet anytime Join Zoom Meeting https://urldefense.us/v3/__https://tntech.zoom.us/j/85895299644?pwd=UVeoehU… <https://urldefense.us/v3/__https://tntech.zoom.us/j/85895299644?pwd=UVeoehU…> Meeting ID: 858 9529 9644 Passcode: mpi4eva --- One tap mobile +13017158592,,85895299644#,,,,*2477715# US (Washington DC) +13052241968,,85895299644#,,,,*2477715# US Join instructions https://urldefense.us/v3/__https://tntech.zoom.us/meetings/85895299644/invi… <https://urldefense.us/v3/__https://tntech.zoom.us/meetings/85895299644/invi…>

1 0

[MPI-Forum] Collective WG meeting today
by Julien.JAEGER＠CEA.FR 25 Feb '26

25 Feb '26

Dear all, MPI Collective WG is meeting today in a bit more than one hour. Please join us to discuss the different topics listed below. Tony Skjellum (TNTECH) is inviting you to a scheduled Zoom meeting. Topic: MPI WG Collective (Persistent too) Time: This is a recurring meeting Meet anytime Join Zoom Meeting https://urldefense.us/v3/__https://tntech.zoom.us/j/85895299644?pwd=UVeoehU… Meeting ID: 858 9529 9644 Passcode: mpi4eva --- One tap mobile +13017158592,,85895299644#,,,,*2477715# US (Washington DC) +13052241968,,85895299644#,,,,*2477715# US Join instructions https://urldefense.us/v3/__https://tntech.zoom.us/meetings/85895299644/invi… -------------------------------- AGENDA - FEB 25, 2026 1. Reconstitution - subsume partitioned, persistent, point-to-point and collective discussions 2. Items for good of the order form participants not covered below 3. Align with hybrid/streams work in pt2pt scope BIG PICTURE AGENDA FOR GROUP, NOT EVERYTHING FOR FIRST MEETING. 1. Discuss partitioned collective status and plans * Prior papers * Issues with synchronizing collectives * Prototypes * New ideas with buffer management * Path to standardization 2. Corollary issues with persistent collectives * Non-blocking init/free - not allowed 3. Related issue - handoff of control of user state * Handling the registration and registration of index vectors and data pointers * C++ way * C (and Fortran) way with registration callbacks at free * Removing ambiguity in the “zero time use case” and whenever a free is allowed to lazy in MPI that holds user state (so it can be released before MPI_Finalize). 1. Managing unexpected state in partitioned (and maybe persistent) operations * Elimination of PBUF_PREPARE in all partitioned considerations * Considering to “generalize” point-to-point partitioned to enable a) persistence better; b) buffer pools 1. Streaming collectives 2. Other things we “owe” short term * MPI_Parrived_any ticket — very useful with current partitioned point-to-point 3. Document enhancements * Revamping the collective chapter front-to-back (ask Puri if he had some input on prior tries at this?) * Making uniform the presentation of blocking, nonblocking, persistent, and future (partitioned) collective * Front matter needs a look for terms, definitions… --- Running minutes will be captured and shared. Recording whenever possible :-)

1 0

Re: [mpiwg-coll] Discussing Semantics and Synchronization Requirements for Neighbourhood Collectives
by Christoph Niethammer 27 Mar '25

27 Mar '25

1 0

Re: [mpiwg-coll] [mpiwg-hw-topology] Discussing Semantics and Synchronization Requirements for Neighbourhood Collectives
by Anthony Skjellum 18 Mar '25

18 Mar '25

FYI to Christoph and Julien, I think that alltoallv/w operations with zero counts are deserving special consideration too … why would we require a barrier ? ; allowing a barrier is acceptable … with neighbors, there is little conceptual difference from zero counts for some sources and sinks, so we must be consistent. The term synchronizing collective means — must Or — may ?? That is confusing. Clearly barrier must… what about persistent allreduce implemented via 1-sided inside … is there for sure a barrier effect ? I am worried about this detail to define partitioned collectives too. Tony Anthony Skjellum, PhD skjellum(a)gmail.com Cell: +1-205-807-4968 On Tue, Mar 18, 2025 at 9:15 AM Christoph Niethammer via mpiwg-hw-topology < mpiwg-hw-topology(a)lists.mpi-forum.org> wrote: > Hello everyone, During the preparation for the FRM, a question was raised > regarding whether and how the semantics of collectives apply to > neighborhood collectives. [1] This question stemmed from remark 18) in the > summary table in Appendix A. 2, > ZjQcmQRYFpfptBannerStart > This Message Is From an External Sender > This message came from outside your organization. > > ZjQcmQRYFpfptBannerEnd > > Hello everyone, > > During the preparation for the FRM, a question was raised regarding > whether and how the semantics of collectives apply to neighborhood collectives. [1] > This question stemmed from remark 18) in the summary table in Appendix A.2, > which was used to apply the semantics of collectives to neighbourhood > collectives at this place. > > The discussions expanded during the March meeting to the question of whether remark 18, > stating that "all-to-all collectives *must* synchronize (with some restrictions)", > is correct. [2] > > I would like to invite you all to discuss these two issues in more detail during a > Topologies WG meeting on March 27th at 15:00, as part of our regular WG slot. > > Best regards > Christoph Niethammer > > [1] https://urldefense.us/v3/__https://github.com/mpi-forum/mpi-issues/issues/8… > [2] https://urldefense.us/v3/__https://github.com/mpi-forum/mpi-issues/issues/9… > > -- > > Dr.-Ing. Christoph Niethammer > High Performance Computing Center Stuttgart (HLRS)Nobelstrasse 19 > 70569 Stuttgart <https://urldefense.us/v3/__https://www.google.com/maps/search/Nobelstrasse*… > > > Tel: ++49(0)711-685-87203 > email: [email protected]://urldefense.us/v3/__https://www.hlrs.de/people/christoph-niethammer__;!!G_uCfscf7eWS!edo6G3QE7JbjtZTTusbpg5yIgnpGmlkZf6PJXEhK0BIFo_p25xzgJ-iVjVZhEuL7z3exLA62R0MjXhpEVDjC5QAigTtN4tiyVz0SIxo$ > > _______________________________________________ > mpiwg-hw-topology mailing list > mpiwg-hw-topology(a)lists.mpi-forum.org > https://urldefense.us/v3/__https://lists.mpi-forum.org/mailman/listinfo/mpi… >

1 0

Re: [mpiwg-coll] [mpiwg-hw-topology] Discussing Semantics and Synchronization Requirements for Neighbourhood Collectives
by Julien.JAEGER＠CEA.FR 18 Mar '25

18 Mar '25

Hello Christoph, it is indeed a good question. I will also take a look at the table w.r.t. inter-communicators. It may have similar caveats for collectives with inter-communicators. Best regards, Julien J. ________________________________ De : mpiwg-hw-topology <mpiwg-hw-topology-bounces(a)lists.mpi-forum.org> de la part de Christoph Niethammer via mpiwg-hw-topology <mpiwg-hw-topology(a)lists.mpi-forum.org> Envoyé : mardi 18 mars 2025 09:15:23 À : mpiwg-hw-topology; mpiwg-semantic-terms(a)lists.mpi-forum.org; MPI-3 Collective Subgroup Discussions Cc : Martin Schulz Objet : [mpiwg-hw-topology] Discussing Semantics and Synchronization Requirements for Neighbourhood Collectives Hello everyone, During the preparation for the FRM, a question was raised regarding whether and how the semantics of collectives apply to neighborhood collectives. [1] This question stemmed from remark 18) in the summary table in Appendix A. 2, ZjQcmQRYFpfptBannerStart This Message Is From an External Sender This message came from outside your organization. ZjQcmQRYFpfptBannerEnd Hello everyone, During the preparation for the FRM, a question was raised regarding whether and how the semantics of collectives apply to neighborhood collectives. [1] This question stemmed from remark 18) in the summary table in Appendix A.2, which was used to apply the semantics of collectives to neighbourhood collectives at this place. The discussions expanded during the March meeting to the question of whether remark 18, stating that "all-to-all collectives *must* synchronize (with some restrictions)", is correct. [2] I would like to invite you all to discuss these two issues in more detail during a Topologies WG meeting on March 27th at 15:00, as part of our regular WG slot. Best regards Christoph Niethammer [1] https://urldefense.us/v3/__https://github.com/mpi-forum/mpi-issues/issues/8… [2] https://urldefense.us/v3/__https://github.com/mpi-forum/mpi-issues/issues/9… -- Dr.-Ing. Christoph Niethammer High Performance Computing Center Stuttgart (HLRS) Nobelstrasse 19 70569 Stuttgart Tel: ++49(0)711-685-87203 email: christoph.niethammer(a)hlrs.de https://urldefense.us/v3/__https://www.hlrs.de/people/christoph-niethammer_…

1 0

Re: [mpiwg-coll] Discussing Semantics and Synchronization Requirements for Neighbourhood Collectives
by Michael Knobloch 18 Mar '25

18 Mar '25

1 0

Discussing Semantics and Synchronization Requirements for Neighbourhood Collectives
by Christoph Niethammer 18 Mar '25

18 Mar '25

1 0