Due to their complex architecture, the radiation sensitivity of general-purpose processors is intrinsically linked to the software that is executed. In order to obtain a consistent assessment of the fault tolerance during radiation experiments, it is necessary to have software representative of the final application. For this purpose, we are developing a methodology based on the application profiling tool that allows us to have a detailed view of the use of hardware resources according to the software profile. This analysis flow has been applied on a reduced instruction set computer (RISC-V) softcore for different applications allowing us to formulate a link between the fault tolerance and the software profile.