Title :
Efficient VLSI layouts of hypercubic networks
Author :
Yeh, Chi-Hsiang ; Varvarigos, Emmanouel A. ; Parhami, Behrooz
Author_Institution :
Dept. of Electr. & Comput. Eng., California Univ., Santa Barbara, CA, USA
Abstract :
In this paper we present efficient VLSI layouts of several hypercubic networks. We show that an N-node hypercube and an N-node cube-connected cycles (CCC) graph can be laid out in 4N2/9+o(N2) and 4N2/(9 log2 2N)+o(N2/log2 N) areas, respectively, both of which are optimal within a factor of 1.7¯+o(1). We introduce the multilayer grid model, and present efficient layouts of hypercubes that use more than 2 layers of wires. We derive efficient layouts for butterfly networks, generalized hypercubes, hierarchical swapped networks, and indirect swapped networks, that are optimal within a factor of 1+o(1). We also present efficient layouts for folded hypercubes, reduced hypercubes, recursive hierarchical swapped networks, and enhanced-cubes, which are the best results reported for these networks thus far
Keywords :
VLSI; hypercube networks; integrated circuit layout; N-node hypercube; VLSI layouts; butterfly networks; cube-connected cycles; enhanced-cubes; folded hypercubes; generalized hypercubes; hierarchical swapped networks; hypercubic networks; indirect swapped networks; multilayer grid model; recursive hierarchical swapped networks; reduced hypercubes; Assembly; Costs; Dairy products; Hypercubes; Multiprocessor interconnection networks; Nonhomogeneous media; Parallel architectures; Propagation delay; Very large scale integration; Wire;
Conference_Titel :
Frontiers of Massively Parallel Computation, 1999. Frontiers '99. The Seventh Symposium on the
Conference_Location :
Annapolis, MD
Print_ISBN :
0-7695-0087-0
DOI :
10.1109/FMPC.1999.750589