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 a method to improve simulation time through handling the structure of the model. This operation consists in partitioning the graph models based on several criteria. In this work, we use the DEVS formalism which is a discrete event formalism with a modular and hierarchical structure of models. To improve simulation time, we use partitioning methods. We will present the partitioning method chosen to achieve this division and quantify the resulting time savings. Many tests are performed from graphs with different sizes and shapes.