Experiences With Multi-Core, HPC Clusters, and GPU Programming
by Richard Moore
Thu, Oct 01, 2009
Room:
Time:
0:00
Based on our experience at 3M speeding up numerical processing algorithms for image processing and ray tracing we will cover topics related to parallel programming in C/C++ on diverse platforms. By looking at examples using threads, MPI, and OpenMP on multi-core workstations, high performance computing clusters, and Nvidia GPUs we will explore some of the issues with programming multiple cores and the performance gains we have achieved. All the examples will be with Visual Studio 2008 C++ and Windows.
About the Author
Richard Moore is a Staff Scientist at 3M Company working in the corporate research Software, Electronics, and Mechanical Systems Laboratory (SEMS). Richard has a Master's of Science in Electrical Engineering from the University of Nevada Reno. His work experience over the last 29 years ranges from microprocessor design to leading teams in the development of commercial software applications and systems. Areas of specialty and interest include image processing, graphics, printing, color, parallel algorithm development, high speed computing architectures among many others.