37 - NHR PerfLab Seminar 2022-10-18: New Features in MPI 4.0/ClipID:45456 previous clip next clip

Recording date 2022-10-18
* Automated closed captions generated with Whisper Open AI

Via

Free

Language

English

Organisational Unit

Zentrum für Nationales Hochleistungsrechnen Erlangen (NHR@FAU)

Producer

Zentrum für Nationales Hochleistungsrechnen Erlangen (NHR@FAU)

Speaker: Dr. Rolf Rabenseifner, High Performance Computing Center Stuttgart (HLRS)

Slides: https://hpc.fau.de/files/2022/10/mpi_4.0_rab_NHR-talk-2022-10-18_v04.pdf

Date and time: Tuesday, October 18, 2:00 p.m. – 3:00 p.m.

Abstract: MPI-4.0 is a major update of the Message Passing Interface standard. It includes significant new functionality. The largest changes are the addition of large-count versions of many routines, persistent collectives, partitioned communications for hybrid MPI+OpenMP, the Sessions Model as an alternative way to initialize MPI for library developers, new ways for the hardware-based split of communicators, clarifications of the neighbor communication, application info assertions, and improvements to the definitions of error handling. In addition, there are a number of smaller improvements and corrections including some semantic changes and warnings. An outlook on MPI-4.1 and 5.0 includes a discussion of the progress rules of MPI and errata on the MPI shared memory interface. The talk wants to answer questions like “What can I do with it?”, “When is this important?”, and “What’s behind?”.

Speaker's Bio: Rolf Rabenseifner studied mathematics and physics at the University of Stuttgart. Since 1984, he has been working at the High-Performance Computing-Center Stuttgart (HLRS). In his dissertation he developed a controlled logical clock as global time for trace-based profiling of parallel and distributed applications. Since 1996, he is a member of the MPI-2 Forum and since Dec. 2007, he was in the steering committee of the MPI-3 Forum and responsible for the MPI-2.1 version of the standard. In workshops and summer schools he teaches parallel programming models at many universities and labs in Germany, Austria, and Switzerland.

See https://hpc.fau.de/research/nhr-perflab-seminar-series/ for past and upcoming NHR PerfLab seminar talks.

More clips in this category "Friedrich-Alexander-Universität Erlangen-Nürnberg Zentralbereich"

2023-03-02
IdM-login
protected  
2023-03-02
IdM-login
protected  
2023-03-13
IdM-login
protected