Known and Potential High Performance Computing Applications in Computer Graphics and Visualization

Abstract

This paper describes past, current, and future applications in computer graphics and scientific visualization which utilize high performance computers. First, a historical perspective will be presented which gives a flavor, though not an all inclusive listing, for computer graphics and visualization implementations for high performance computers. Next, the current state of the field is described. Finally, some observations about applications for computer graphics and visualization in high performance computing are presented. One thing to note is the variety of applications which have differing emphasis in terms of algorithmic development. Computer rendering algorithms which produce photorealistic images may take from minutes to hours to render an image but the goal is producing a high degree of realism as fast as possible. Whereas, scientific visualization algorithms typically strive for real-time rendering of images often sacrificing image quality for rendering speed (i.e. gouraud shading vs ray-tracing). Obviously, this leads to a dichotomy for these two differing application areas. One needs to be cognizant of the underlying goals for a particular algorithm when comparing implementations.