Monday-Tuesday: Introduction to Hybrid Programming in HPC (open to everybody)
Most HPC systems are clusters of shared memory nodes. To use such systems efficiently both memory consumption and communication time has to be optimized. Therefore, hybrid programming may combine the distributed memory parallelization on the node interconnect (e.g., with MPI) with the shared memory parallelization inside of each node (e.g., with OpenMP or MPI-3.0 shared memory). This course analyzes the strengths and weaknesses of several parallel programming models on clusters of SMP nodes. Multi-socket-multi-core systems in highly parallel environments are given special consideration. MPI-3.0 has introduced a new shared memory programming interface, which can be combined with inter-node MPI communication. It can be used for direct neighbor accesses similar to OpenMP or for direct halo copies, and enables new hybrid programming models. These models are compared with various hybrid MPI+OpenMP approaches and pure MPI. Numerous case studies and micro-benchmarks demonstrate the performance-related aspects of hybrid programming.
Hands-on sessions are included on both days. Tools for hybrid programming such as thread/process placement support and performance analysis are presented in a "how-to" section. This course provides scientific training in Computational Science, and in addition, the scientific exchange of the participants among themselves. This course is organized by HLRS, in cooperation with RRZE and VSC (Vienna Scientific Cluster).
Target audience: Scientists developing HPC applications with MPI.
Although these two days are independent of any specific hardware, these "hybrid days" may be a good basis for the challenges when porting your application to Hawk, the new AMD Rome based HPE cluster at HLRS with 128 cores per node.
Wednesday-Friday: Porting and Optimization Workshop (only for customers of HLRS)
The Wednesday morning is dedicated to introductions to the new system (hardware characteristics, batch system, compilers, MPI, ...).
In this workshop, users can port their applications to HLRS' new AMD-based HPE Apollo supercomputer "Hawk" (to be installed in early 2020) assisted by HLRS and HPE staff. For the course, Hawk will not be available, but you can start porting your application on our small Hawk test system.
In order to achieve usable efficiency enhancements, it is important to discuss pros and cons of potential solutions. This, however, requires application as well as machine expertise. Hence, this workshop brings together our users (with their application expertise) and support staff (with their machine expertise).
Target audience: Groups holding a compute time budget to be used on Hawk.
We combined both course parts, because the hybrid days may be a good basis for the challenges when porting your application to Hawk with its 128 cores per node.
HLRS, University of Stuttgart Nobelstraße 19 70569 Stuttgart, Germany Room 0.439 / Rühle Saal Location and nearby accommodations
27. Jan 2020 08:45
31. Jan 2020 15:30
Stuttgart, Germany
Englisch
Fortgeschritten
Paralleles Programmieren
Code-Optimierung
MPI
OpenMP
Zurück zur Liste
1st day (Hybrid programming, part 1)
08:45 Registration 09:00 Welcome 09:05 Motivation 09:15 Introduction 09:45 Programming Models 09:50 - MPI + OpenMP 10:30 Coffee Break 10:50 - continue: MPI + OpenMP 11:40 Practical (how to compile and start) 12:30 Practical (hybrid through OpenMP parallelization) 13:00 Lunch 14:00 Practical (continued) 15:00 Coffee Break 15:20 - Overlapping Communication and Computation 15:40 Practical (taskloops) 16:20 - MPI + OpenMP Conclusions 16:30 - MPI + Accelerators 16:45 Tools 17:00 End of first day 2nd day (Hybrid programming, part 2)
09:00 Programming Models (continued) 09:05 - MPI + MPI-3.0 Shared Memory 09:45 Practical (replicated data) 10:30 Coffee break 10:50 continue: Practical (replicated data) 11:50 - MPI Memory Models and Synchronization 12:30 Lunch 13:30 - Pure MPI 13:50 - Topology Optimization 14:30 Coffee Break 14:50 Practical (application aware Cartesian topology) 15:45 - Topology Optimization (Wrap up) 16:00 Conclusions 16:15 Q & A 16:30 End of second day (course)
PORTING AND OPTIMIZATION PART (Wed-Fri)
Supported by HLRS and HPE specialists, you will learn how to port your application to the new supercomputer system “Hawk”. All categories of bottlenecks (CPU, memory subsystem, communication and I/O) will be addressed, according to the respective requirements.
3rd day (Porting workshop):
08:45 Registration (for new participants) 09:00 Welcome (pdf) and Overview (pdf): recording 09:15 The AMD EPYC Rome processor: pdf and recording 09:45 Network Topology and HPE MPI: pdf and recording 10:15 Module Environment on Hawk: pdf and recording 10:30 Batch System: pdf and recording 10:50 Coffee Break on demand: Performance Tools 13:00 Lunch 14:00 Porting of own applications to Hawk (on the Hawk test system) 17:30 End of third day
4th day (Porting workshop, continued):
09:00-09:15 Introduction 09:15-17:30 Porting of own applications to Hawk (on the Hawk test system) - continued
5th day (Porting workshop, continued):
09:00-15:30 dto.
We recommend to book cancellable tickets and hotels only, so that you can leave as soon as your porting work is done.
The talks on Wednesday morning will also be provided via a video conference system.
Please join the videoconference with your favourite browser on Windows, MacOS, GNU/Linux:
https://conf.dfn.de/webapp/conference/97988854
Important:
Hybrid programming days: Basic MPI and OpenMP knowledge as presented, e.g., in our Training Courses on MPI and OpenMP. For the hands-on sessions you should know Unix/Linux and either C/C++ or Fortran in particular.
Scaling workshop: Your group holds a compute time budget to be used on Hawk.
Learn more about course curricula and content levels.
The course language is English.
Hybrid programming days: See http://tiny.cc/MPIX-HLRS.
Hybrid programming days: Dr. habil. Georg Hager (RRZE/HPC, Uni. Erlangen), Dr. Rolf Rabenseifner (HLRS, Uni. Stuttgart), Dr. Claudia Blaas-Schenner and Dr. Irene Reichl (VSC Team, TU Wien)
Porting and optimization days: HLRS and HPE staff
Registration is now closed.
for registration is Jan. 19, 2020 (extended deadline)
Late registrations after the deadline are still possible but maybe with reduced quality of the handouts.
Members of universities and public research institutes within EU or PRACE member countries, registering only Wednesday - Friday: 0 EUR Students without Diploma/Master: 25 EUR Students with Diploma/Master (PhD students) at German universities: 45 EUR Members of German universities and public research institutes: 45 EUR Members of universities and public research institutes within EU or PRACE member countries: 90 EUR. Members of other universities and public research institutes: 180 EUR Others: 420 EUR
(includes coffee breaks)
see our How to find us page.
HLRS is part of the Gauss Centre for Supercomputing (GCS), which is one of the six PRACE Advanced Training Centres (PATCs) that started in Feb. 2012. HLRS is also member of the Baden-Württemberg initiative bwHPC-C5. This course is provided within the framework of the bwHPC-C5 user support. This course is not part of the PATC curriculum and is not sponsored by the PATC program.
Rolf Rabenseifner phone 0711 685 65530, rabenseifner(at)hlrs.de Lorenzo Zanon phone 0711 685 63824, zanon(at)hlrs.de Björn Dick phone 0711 685 87189, dick(at)hlrs.de
http://www.hlrs.de/training/2020/HY-S
http://www.hlrs.de/training/ and http://www.hlrs.de/training/overview/ (at HLRS)
Januar 13 - 31, 2025
Hybrid Event - Stuttgart, Germany
Januar 21 - 23, 2025
Februar 17 - 21, 2025
März 17 - 21, 2025
Dresden, Germany
März 24 - 28, 2025
