Rendering polygons and computing geodesic offset curves in a parallel processing environment

Research is described which involves computer graphics, computer aided geometric design (CAGD) and parallel processing. The research involves the design and implementation of two software applications on the highly scalable parallel processing machine, the Connection Machine model CM-5, located at the Advanced Computing Laboratory (ACL) of Los Alamos National Laboratory (LANL). A message passing model is used to develop both programs.; One application focuses in the area of computer graphics and involves the high performance rendering of large volumes of polygonal data. A new rendering algorithm is described which is specifically designed for a distributed memory, multiple-instruction multiple-data (MIMD) machine. The other application focuses in the area of CAGD and involves the approximation of geodesic offset curves. A new algorithm is examined which reduces the complexity of the calculations previously required to generate a solution. The design and implementation of a program to compute geodesic offset curves on an IRIS (Silicon Graphics) workstation located at Arizona State University is also developed.; Following a brief introduction, a description of some parallel processing terminology and the CM-5 environment at the ACL is given. The motivation for the development of both applications is then stated and their design and implementation are described. Specifically, the mapping of the graphics pipeline to the multiple, general purpose processors of the CM-5 and the underlying geometry of an iterative scheme for tracing geodesic paths are detailed. The resulting programs are then examined in terms of timing statistics. Timing statistics are illustrated in terms of the number of processors used and the complexity of the input data. Conclusions are then drawn about the success of the implementations and ideas concerning further research are outlined.