Instructors:

Dr. Jan Meinke, Jochen Kreutz, Dominic Eschweiler, Willi Homberg, JSC;

Dr. Daniel Becker, GRS

Contents:

Many-core programming is a very dynamic research area. Many scientific applications have been ported to GPU architectures during the past four years. We will give an introduction to CUDA, OpenCL, and multi-GPU programming using examples of increasing complexity. After introducing the basics the focus will be on optimization and tuning of scientific applications. Topics covered will include:

• CUDA programming model and language extension
• Partitioning and granularity of parallel applications
• Debugging and profiling of CUDA kernels
• Impact of memory access patterns on performance
• Multi-GPU programming using MPI
• OpenCL programming model and API

Prerequisites:

Knowledge in C

Duration:

3 days

Dates:

21 - 23 March 2011, 9:00 - 16:30

Venue:

Ausbildungsraum 1, Jülich Supercomputing Centre, building 16.3, room 021

Registration:

A minimum of 5 participants is required. Registration is necessary until 10 March 2011.
To register, please contact Willi Homberg,