Concurrent with SC09, PGI is offering the tutorial “GPU Programming with the PGI Accelerator Programming Model and PGI CUDA Fortran”. The tutorial is Monday, 16 November, 2009 in Portland, a few blocks from the Oregon Convention Center.
The full day course starts with an overview of GPU programming including helpful tips for those just getting started with GPUs. The tutorial includes segments on programming GPUs using CUDA Fortran and using the PGI Accelerator programming model in Fortran and C99. Using real-world examples to demonstrate concepts, each segment includes an overview along with guidance for evaluating when and where to use each model. The tutorial describes a work flow you can use for porting your code to GPUs. It starts with identifying good candidate sections of your code, illustrates different methods for modifying your code to run on GPUs, and explains how to interpret compiler feedback. The tutorial concludes with a discussion of techniques for optimizing performance.
The tutorial will be presented by Michael Wolfe. The price is $500 US and includes a hosted lunch, an afternoon reception, and a copy of the PGI Accelerator Fortran/C/C++ Workstation compilers and tools with CUDA Fortran. More information is available on the PGI website at www.pgroup.com/support/tutorial_outline.htm.