This paper evaluates different parallel approaches for multipopulation genetic algorithm. These approaches are applied to solve a synchronized and integrated lot sizing and scheduling problem. In this problem, the challenge is to simultaneously determine lot sizing and scheduling for raw materials in tanks and products in lines. First, the parallel algorithms are designed to be executed using a multicore server. The best approach is also executed by duo core computers using MPI. A set of real-world instances found in the literature are solved. Also, a new set of instances is proposed. The speedups improvements are showed as well as the quality of final solutions found.