The work presented in this paper aims at restricting the input parameter values of the semi-analytical model used in GALICS and MOMAF, so as to derive which parameters influence the most the results, e.g., star formation, feedback and halo recycling efficiencies, etc. Our approach is to proceed empirically: we run lots of simulations and derive the correct ranges of values. The computation time needed is so large, that we need to run on a grid of computers. Hence, we model GALICS and MOMAF execution time and output files size, and run the simulation using a grid middleware: DIET. All the complexity of accessing resources, scheduling simulations and managing data is harnessed by DIET and hidden behind a web portal accessible to the users.

One way to access the aggregated power of a collection of heterogeneous machines is to use a grid middleware, such as DIET, GridSolve or NINF. It addresses the problem of monitoring the resources, of handling the submissions of jobs and as an example the inherent transfer of input and output data, in place of the user. In this paper we present how to run cosmological simulations using the RAMSES application along with the DIET middleware. We will describe how to write the corresponding DIET client and server. The remainder of the paper is organized as follows: Section 2 presents the DIET middleware. Section 3 describes the RAMSES cosmological software and simulations, and how to interface it with DIET. We show how to write a client and a server in Section 4. Finally, Section 5 presents the experiments realized on Grid'5000, the French Research Grid, and we conclude in Section 6.