This course explored techniques for optimizing locality of reference, efficient processing of graphics calls through the software/hardware pipeline, efficient use of graphics hardware resources, and software techniques to maximize graphics performance. It reviewed interactions among CPU, accelerator, API, software driver, cache, memory, and bus. Cross-platform PC issues were discussed, including commonality among feature sets, scalability, and choice of graphics API's.
Prerequisites
Experience in graphics programming. Understanding of basic computer architecture (memory hierarchy, con-current execution, etc.) helpful. This course was explicitly oriented toward the programmer.
Topics Covered
The essentials of the graphics software and hardware architecture for the PC platform and the programming techniques required to write the highest-performance graphics applications possible.
Organizers
Michael Cox
MRJ Technology Solutions/NASA Ames Research Center
David Sprague
Intel Corporation
Lecturers
John Danskin
Dynamic Pictures
Rich Ehlers
Evans & Sutherland Computer Corporation
Brian Hook
id Software
Bill Lorensen
General Electric Corporate R&D Center
Gary Tarolli
3Dfx Interactive
