Developing a Cluster Based Parallel Ray Tracer

Parallel ray tracing is an important concept in computer graphics, as it allows for high quality ray traced images to be produced at a faster rate than is possible in single processor ray tracing. Presented in this paper is the conversion of a previously created single processor ray tracer to a parallel ray tracer that is capable of running on a cluster of computers. This paper presents the technical aspects of designing a parallel ray tracer by looking at the theory of transforming ray tracing into a parallel process and by comparing the Condor and Linux based cluster computing technologies. This paper then highlights how the methods available for turning ray tracing into a parallel process can be dependent on the cluster computing technology going to be used