Recursive Diagonal Torus (RDT), a class of interconnection network is propo
sed for massively parallel computers with up to 2(16) nodes. By making the
best use of a recursively structured diagonal mesh (torus) connection, the
RDT has a smaller diameter (e.g., it is II for 2(16) nodes) with a smaller
number of links per node (i.e., 8 links per node) than those of the hypercu
be. A simple routing algorithm, called vector routing, which is near-optima
l and easy to implement is also proposed. Although the congestion on upper
rank tori sometimes degrades the performance under the random traffic, the
RDT provides much better performance than that of a 2D/3D torus in most cas
es and, under hot spot traffic, the RDT provides much better performance th
an that of a 2D/3D/4D torus. The RDT router chip which provides a message m
ulticast for maintaining cache consistency is available. Using the 0.5 mum
BICMOS SOG technology, versatile functions, including hierarchical multicas
ting, combining acknowledge packets, shooting down/restart mechanism, and t
ime-out/setup mechanisms, work at a 600MHz clock rate.