We propose to parallelize an existing implementation of a ray tracer called
Raylab, which is a public domain software.
This sample scene (640x480, reduced here) took approximately 11 minutes to
render on the sequential version.
We have finished learning CC++ and now intend to investigate strategies for
parallelization of the code. We will then convert the program first to C++
(program is currently in C) and then to CC++.
We hope to come close to a speed-up of N. The program will be implemented
on the SP-2 port of CC++.