[en] Following the present tendency of the international community, the French industrial partners are greatly interested in the technical and economical potential performances of the HTGR (High Temperature Gas cooled Reactor) concept that appears as a promising reactor for the future nuclear power applications. However, the next nuclear program related studies will rely on a major part on the intensive use of computational calculations rather than experimental results. Therefore, the codes will have to be reliable and well prepared to calculate a wide range of applications. The computational tools will have to serve as well for conceptual studies and industrial calculations as for best estimate and reference calculations. Taking advantages of the benchmark problems of the HTTR's start-up core physics experiments initially proposed by JAERI through the IAEA in a Coordinated Research Program, CEA has performed calculations in order to validate and qualify the codes that will serve to evaluate the future HTGR generation. For the analysis, two different calculation schemes based on a deterministic or a probabilistic approach are used. It turns out that all calculation results obtained for the fully loaded core configuration fit well each other and with the experiment, considering the experimental uncertainties. As far as the diffusion calculations are concerned, the number of fuel columns needed to achieve criticality increases by about 7 or 8 in comparison with the results obtained with a simplified Transport - Diffusion calculation scheme (homogenized fuel elements and no streaming effect taken into account in the diffusion calculations). (author)