If you are running VM's on the nMP does (or can) the VM take advantage of the second GPU? From what I've read there are few applications that can actually take advantage of the second GPU - but can a VM use the second GPU independent of the first? or is it a shared resource?