In this paper, we implement ACO algorithms on a PC which has 4 GTX 480 GPUs. We implement two types of ACO models; the island model, and the master/slave model. When we compare the island model and the master/slave model, the island model shows promising speedup values on class (iv) QAP instances. On the other hand, the master/slave model showed promising speedup values on both classes (i) and (iv) with large-size QAP instances.
