A Faster SA Method TOSA for Global Placement

In this paper, we propose a new ideas of the Simulated Annealing method to improve the execution time. SA method is an effective optimization algorithm for combinatorial problems such as a placement design for VLSI Chip, but a deficiency is the long execution time. Therefore, to improve the execution time of SA algorithm is widespread concern. In this paper, we propose a new idea TOSA that traces optimal elements to reduce ineffective select in the simulated Annealing process. We applied TOSA to the standard-cell global placement problems, premise the quality of solution and achieved more than 40 % faster time at most than those of simulated annealing.