A Multiprocessor Architecture for Generating Fractal Surfaces

Fractal surfaces have recently been shown to be a useful model for generating images of terrain in computer graphicS. Unfortunately, the generation of fractal images is very costly in CPU time. A multiprocessor architecture is described which takes advantage of the parallelism inherent in fractals to speed the generation of images. The performance of the processing array is analyzed along with the suitability of implementation in VLSI.