I could think of an application in which there would be minimal disk interaction and just lots of network action and in-memory processing so hypothetically it a

I could think of an application in which there would be minimal disk interaction and just lots of network action and in-memory processing so hypothetically it all depends on how the CPU handles parallelization through its cores.