GPGPU stands for General-Purpose Computation Using Graphics Hardware: The raw computational power of commodity graphics processors (GPUs) exceeds CPU peak performance by at least an order of magnitude, and GPU performance increases at a significantly faster rate than CPU performance. Coupled with recent advances in programmability and floating point support, GPUs have evolved into a highly attractive platform for computationally demanding tasks in a wide variety of application domains: GPUs are are no longer restricted to the graphics-specific rendering tasks for which they were originally designed, instead, they are now capable coprocessors for general purpose computations (GPGPU).
This event took place in conjunction with the International Conference on Computational Science 2006 (ICCS 2006), Reading, UK, May 2006.
A six hour tutorial session presenting basic and advanced GPGPU techniques and applications on an introductory level took place in addition to the main workshop. Course notes and materials are available on these pages.
GPGPU Workshop: Methods, Algorithms and Applications
The workshop comprised a total of eight paper contributions, covering a wide variety of GPGPU applications. Abstracts and citations are available on these pages, and the full proceedings have been published in the LNCS series, Springer.
Please refer to the workshop web page at http://www.mathematik.uni-dortmund.de/~goeddeke/iccs for additional information or contact Dominik Göddeke (firstname.lastname@example.org).