Search

Quadratic/Analytical placement methods have been widely used in the latest IC design process. This kind of placement needs a powerful detailed placement method for large-scale mixed macro and standard cell placement. An efficient and effective multi-stage iterative detailed placement (MSIP) algorithm is proposed in this paper. It combines a better initial placement based on combinatorial optimization method with a deterministic local search for post optimization. Various strategies are used for saving computation time. Experimental results show that it can get an average of 22% wire length improvement comparing to PAFLO [6] in comparable runtime. [ABSTRACT FROM AUTHOR]