Parallel Programming in Computational Engineering and Science 2011
HPC Seminar and Workshop
Monday, March 21, - Friday, March 25, 2011
This event which continued the tradition of previous annual week-long events taking place in Aachen every spring since 2001 was a complete success!!.
Beginning with a series of overview presentations on Monday afternoon, we are very happy to announce the upcoming RWTH Compute Cluster.
Agenda for monday special see below:
Seminar and Workshop
Throughout the week, we will cover serial and parallel programming using OpenMP and MPI in Fortran and C / C++ as well as performance tuning addressing both, Linux and Windows platforms. Due to the positive experience last year, we are happy to present a renowned speaker to give an introduction into GPGPU architectures and programming.
- Dieter an Mey (RWTH Aachen University)
- Thomas Warschko (Bull)
- Herbert Cornelius (Intel)
- Jean-Pierre Panziera (Bull)
- Christian Bischof (RWTH Aachen University)
- Ruud van der Pas (Oracle)
- Felix Wolf (German Research School for Simulation Science)
- Michael. Wolfe (PGI)
- HPC Team of the RWTH Aachen University
Attendees should be comfortable with C/C++ or Fortran programming and interested in learning more about the technical details of application tuning and parallelization. The presentations will be given in English language. Please fill out the remarks field during the registration, if you want to use Windows or Linux.
There will be the sponsored social event on Wednesday, March 23 at 7pm in Restaurant Kazan in Annastraße in nice atmosphere with tasty food. Please feel free to participate in the social dinner on Wednesday, even if only visit one of the workshop days.
Agenda and Registration
Download Agenda (last update: March 16)
Registration (separate for each day!)
Deadline on Monday March 21
Monday, March 21
14:00 - 17:30
Announcement of the upcoming RWTH HPC-Cluster
J.-P. Panziera, T. Warschko (Bull), H. Cornelius (Intel), F. Wolf (GRS), R. van der Pas (Oracle) and C. Bischof, D. an Mey (RWTH)
Tuesday, March 22
9:00 - 17:30
|Serial Programming, Tuning, Debugging, Processor Architectures|
After an introduction into parallel computer architecture, we will focus on performance tuning issues and debugging. Furthermore basic introductions into the programming development environments on Windows and Linux will be presented. For those who feel comfortable with one platform, this will also be a nice opportunity to get hands on the other one.
R.v.d. Pas (Oracle)
Wednesday, March 23
9:00 - 17:30
|Shared Memory Programming with OpenMP|
OpenMP is a widely used approach for programming shared memory architectures, which is supported by most of the compilers nowadays. We will cover the basics of the programming paradigm as well as some advanced topics, such as programming NUMA machines or clusters coherently coupled with vSMP software from ScaleMP.
RWTH HPC Team
R.v.d. Pas (Oracle)
Social Dinner Restaurant Kazan
Thursday, March 24
Message Passing with MPI
The Message Passing Interface (MPI) is the de facto standard for programming large HPC-Clusters. We will introduce the basic concepts, give an overview of advanced features and an introduction into a selection of performace tools.
Friday, March 25
9:00 - 15:30
We study GPU programming as a parallel programming skill. We briefly introduce GPU architectures and why GPU programming differs from multicore parallel programming. CUDA is NVIDIA‘s architecture for executing highly parallel applications on GPUs; we introduce NVIDIA‘s CUDA C extensions and PGI‘s CUDA Fortran extensions, showing basic concepts, performance measurement, and tuning. Then we look at the PGI Accelerator Model, a directive-based approach to programming GPUs.
The exercise session in the afternoon will cover the basic concepts of GPGPU programming. For the first approaches, we will provide CUDA-capable laptops. In addition, we will give a practical introduction to the batch usage of the available high-end GPGPUs belonging to the RWTH compute cluster. At the end of Q2 2011 a GPU-Cluster comprising approx. 50 NVIDIA Fermi graphics cards will be established to power the new CAVE of the RWTH Virtual Reality Group and for high performance GPGPU computing.
M. Wolfe (PGI)
Michael Wolfe has worked in compilers in academia and industry for over 30 years. He has been a compiler engineer at The Portland Group, Inc. (PGI) since 1996, where he currently leads development of GPU programming.
Programming Envirement and Serial Tuning