Standard Cell Placement using Iterative & Constructive Heuristics for Multi-Objective Optimization

Focus in the present rests on optimization of multi-objective standard cell placement for maximizing the speed and minimizing power and interconnect wire-length with cell-width as a constraint. It incorporates fuzzy cost function rules for designing of multi-objectives to integrate the cost of above defined objectives. A significant improvement is reported when the iterative-constructive standard cell-placer combines a course grain timing-driven partitioning followed by adaptive time driven construction of rows as well as selection of driver size. Iterative heuristic algorithms have been developed on Genetic algorithm. These techniques are applied to standard cell ami libraries 20, 33 and 120. 20% improvement in results on average running; almost twice faster than the commercial tools; and reduction of power by 12% is reported for the same timing and wire-length.