OPTICAL BINARY DE BRUIJN NETWORKS FOR MASSIVELY-PARALLEL COMPUTING - DESIGN METHODOLOGY AND FEASIBILITY STUDY

The interconnection network structure can be the deciding and limiting factor in the cost and the performance of parallel computers. One of the most popular point-to-point interconnection networks for parallel computers today is the hypercube. The regularity, logarithmic diameter , symmetry, high connectivity, fault tolerance, simple routing, and re configurability (easy embedding of other network topologies) of the hy percube make it a very attractive choice for parallel computers. Unfor tunately the hypercube possesses a major drawback, which is the comple xity of its node structure: the number of links per node increases as the network grows in size. As an alternative to the hypercube, the bin ary de Bruijn (BdB) network has recently received much attention. The BdB not only provides a logarithmic diameter, fault tolerance, and sim ple routing but also requires fewer links than the hypercube for the s ame network size. Additionally, a major advantage of the BdB network i s a constant node degree: the number of edges per node is independent of the network size. This makes it very desirable for large-scale para llel systems. However, because of its asymmetrical nature and global c onnectivity, it poses a major challenge for VLSI technology. Optics, o wing to its three-dimensional and global-connectivity nature, seems to be very suitable for implementing BdB networks. We present an impleme ntation methodology for optical BdB networks. The distinctive feature of the proposed implementation methodology is partitionability of the network into a few primitive operations that can be implemented effici ently. We further show feasibility of the presented design methodology by proposing an optical implementation of the BdB network.