With the emergence of parallel computational infrastructures at low cost, reducing simulation time becomes again an issue of the research community in modeling and simulation. This paper presents an approach to improve time of discrete event simulations. For that, the PDEVS formalism is coupled to a partitioning method in order to parallelize the graph of models. We’ll present the graph partitioning method to realize this cutting and quantify the resulting time savings of parallel implementation. This article highlights the importance of considering the dynamic of the model when partitioning to improve performances. Many tests are performed from graphs with different sizes and shapes on several hardware architectures.