Part I: Introduction, Parallel Computing Architectures, Serial Tuning – Monday, March 19

After an introduction to the principles of today’s parallel computing architectures, the configuration of the new components of the RWTH Compute Cluster delivered by the company Bull will be explained. As good serial performance is the basis for good parallel performance, we cover serial tuning before introducing parallelization paradigms.

– During your registration, please indicate whether you plan to take part in the introduction to MS Visual Studio at Monday, 11 am  in the „remarks“ field.

– During your registration, please indicate whether you plan to use Linux and/or Windows for your hands-on exercises in the „remarks“ field.

Part II: Message Passing with MPI – Tuesday, March 20

The Message Passing Interface (MPI) is the de-facto standard for programming large HPC Clusters. We will introduce the basic concepts and give an overview of some advanced features. Furthermore, we will introduce the TotalView debugger and a selection of performance tools. We will also cover hybrid parallelization, i.e. the combination of MPI and shared memory programming. Hybrid parallelization is gaining popularity as the number of cores per cluster node is growing.

– During your registration, please indicate whether you plan to use Linux and/or Windows for your hands-on exercises in the „remarks“ field.

Part III: Shared Memory Programming with OpenMP – Wednesday, March 21

– During your registration, please indicate whether you plan to take part in the social event on Wednesday 7 pm  in the „remarks“ field.

– During your registration, please indicate whether you plan to use Linux and/or Windows for your hands-on exercises in the „remarks“ field.

Part IV: GPGPU Programming – Thursday, March 22

We will study GPU programming as a parallel programming skill. We will briefly introduce GPU architectures and explain why GPU programming differs from multicore parallel programming. CUDA is NVIDIA‘s architecture for executing highly parallel applications on their GPUs; we will introduce NVIDIA‘s CUDA C extensions in comparison with the OpenCL standard, showing basic concepts, performance measurement and tuning. Then, we will look at the PGI Accelerator Model, a directive-based approach to program GPUs and give a glimpse into the upcoming OpenACC programming paradigm for accelerators.

Part V: Lab Exercises, Tune your own Code – Friday, March 23

At the end of a week full of presentations and hands-on sessions, we would like to give you the opportunity to dive into some more practical details, to continue working on lab exercises or to get started with porting or tuning your own codes. Don’t put your expectations too high though, as time for working on large codes is limited. You will profit more from this opportunity the better you are prepared.

– During your registration, please indicate whether you plan to use Linux and/or Windows for your hands-on exercises in the „remarks“ field.

– During your registration, please indicate whether you plan to bring your own code in the „remarks“ field.

Speakers

• Thomas Warschko and Oliver Fortmeier, Bull
• Bernd Dammann, DTU, Lyngby, Denmark
• Members of the HPC Team of the RWTH Center for Computing and Communication

Course Materials

Costs

• The previous PPCES 2011 event.
• The HPC Web site of the RWTH Center for Computing and Communication
• The RWTH Compute Cluster User’s Guide

Location