For as long as we can remember, the desktop-to-mobile GPU conversion math was pretty simple: The big chip goes in the desktop, and the smaller version goes in the laptop. That's what Nvidia has been ...