Bipartitioning circuits using TABU search

Abstract

VLSI circuit partitioning is an extensively studied problem. Various iterative improvement based heuristics have been, proposed for obtaining reasonably good solutions. In this paper we explore the applicability of TABU search for partitioning of electrical circuits. We briefly describe the TABU search and its application to circuit or graph partitioning. We have generated experimental results on a variety of standard benchmark circuits. Our results match in quality, and at times improve, the tightest known results in the partitioning literature. We have shown extensive comparison with at least six very well known methods for circuit partitioning.