Difference between sub-modeling and make rigid

Hi,  can

I saw a couple of simulation new sub-modeling feature. From what I understand, under certain conditions (explained in the solidworks 2013 whats new guide) you can use a submodel where you will use a denser mesh.     I imagine the software will tranfer the freebody forces from the whole model in order to represent the correct tensional state.

Now my question is: what's the difference between making rigid the parts you're not interested into and creating a sub-model of the part you're interested into ?

I imagine the difference will be huge when the whole structure deforms a lot, but when load and displacements are concentrated in the part of interest I can't figure out what could be the advantages of sub-modeling over the make rigid command.

Thanks,

Alex

SolidworksSimulation