I don't know if this is a ridiculous concept, but has there been any research or talk regarding the use of Tesla GPUs as VM cores? Basically, I am talking about expanding the capacity of a host to support more VMs that rely on Tesla core processing.
↧