Stress value at extreme fiber: half of theoretical value

I modeled a beam (fully restrained on both ends) via a solid meshing (hence not considered as a physical beam by the software, but as a solid). When I run the simulation, I get the right value for the total displacement (wL^4/(384EI)). However, I don’t get the correct value for the stress on the extreme fiber of the beam, at mid-span. Indeed, with my hand calculations, I get twice the stress that solidworks gives me. I used the formula (stress = M*y/I) to do my hand calculations. Solidworks gives me the exact same result, but exactly divided by two (at two decimal units). I did several attempts (changed the length of the beam, the size of the mesh, the material…), but the stress is still exactly (two decimal units) half of the result I should obtain, despite the fact that the deflection is still correct (which is not function of the mesh used). I don’t know what causes this problem…

SolidworksSimulation