The advent of programmable graphics hardware has opened up a new paradigm in high performance computing. Performance hungry applications can now tap into these massively parallel processors and achieve order of magnitude speedups. This course provides an introduction to this emerging technology. The course will combine lectures on architectural concepts, with hands on software programming exercises to help students learn more about this field and apply it to their application areas.

Topics Covered:

• GPU Architecture and SIMD fundamentals
• Programmable Graphics Pipeline
• Overview of GPU Programming Languages - GLSL, HLSL
• Vendor-specific GPU Compute APIs (nVidia and ATI)
• Detailed Tutorial on any one GPU Compute API
• Compute Kernel development, debugging and tuning
• Programming Assignment
• Performance Optimizations

Target Audience:
This course is intended for software professionals and students who are responsible for or interested in developing high performance computing applications.

Prerequisites:
Experience with C/C++ programming is a must and will be assumed. Familiarity with UNIX or Windows operating systems and development environments is necessary. Understanding of basic concepts in computer architecture will be helpful.