A Novel Ant Colony Optimization Based Temperature-Aware Floorplanning Algorithm

In this paper, an ant colony optimization based temperature-aware floorplanning algorithm is proposed. The object of temperature-aware floorplanning is to decrease the temperature gradient and the area occupancy simultaneously. First, the problem is presented based on Ant Colony Optimization (ACO). Then, with multi-objective optimization, two methods are adopted for decreasing the temperature difference and the area occupancy for the first time for MCNC benchmark Circuits. Simulation Results show that the DeltaTmax can be at most optimized to 7.8degC or 14.6deg C when the DeltaTmax for initial floorplan is 27.4deg C, while the best area occupancy is 87.9% or 89.7%. Compared to Simulated Annealing (SA) algorithm, our results show great improvement in speed with little area latency.