We propose a new parallel ray-tracing hardware architecture in which p
rocessors are connected as a ring. Most parallel ray-tracing algorithm
s subdivide the whole object space into subregions; A processor handle
s only rays entering the subregion assigned to it. Here we assign each
processor objects that are spread over the whole object space. The pr
ocessors trace rays on their own objects. The respective partial resul
ts are combined to form the final image. This scheme is especially sui
table for synthesizing animated sequences because objects need not be
reallocated for every frame. Preliminary results show a speed-up facto
r almost linearly proportional to the number of processors.