Issue No. 10 - October (1994 vol. 5)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/71.313118
<p>We consider primarily the simulation of large networks by smaller ones-an importantconsideration, because interconnection networks are typically of a fixed size, and yetapplications may employ networks of a larger size. Current research (Dingle andSudborough, 1993) describes methods to simulate common data structures and networkarchitectures on the pyramid. However, these simulations assume that the pyramid growswith the size of the network or data structure. Because unbounded growth is notfeasible, we address the issue of mapping several points of the guest data structure ornetwork to a single host processor. We determine how a small pyramid may efficientlysimulate the computation of a larger pyramid as well as that of tree networks.</p>
Index Termsdata structures; graph theory; resource allocation; multiprocessor interconnectionnetworks; trees (mathematics); optimisation; pyramid network mapping; large networksimulation; small networks; interconnection networks; data structures; networkarchitectures; unbounded growth; single host processor; tree networks; graphembeddings; parallel networks; multiprocessors; communication cost; load balancing;simulation modelling
A. Dingle and I. Sudborough, "Efficient Mappings of Pyramid Networks," in IEEE Transactions on Parallel & Distributed Systems, vol. 5, no. , pp. 1009-1017, 1994.