To solve practically large-scale mixed integer programs that appear in various problem-solving in process systems, e.g., logistics, facility location planning, scheduling, etc., we here propose a novel parallel computing algorithm for a hybrid genetic algorithm, HybGA. Since the solution of such mixed integer programs generally requires a large amount of calculation, we attempted to improve the solution efficiency through parallel computing. We first examined the conventional parallel genetic algorithms known as master-slave (MS) parallel GA and Island GA, and showed that the equivalent structure between the HybGA and MS parallel GA is suitable to implement the MS parallel HybGA on the PC cluster machines through a PVM (parallel virtual machine). In addition, we found that equally divided asynchronous distribution is optimal as a dynamic distribution method between master and slave machines. Finally, taking the location problem of waste disposal sites, we verified the effectiveness of the proposed algorithm through numerical experiments carried out under various problem sizes.