A couple weeks ago I wrote about NVIDIA’s latest announcements around Tesla and their personal supercomputer spec. In that post I said this:
“Here NVIDIA is trying to use a brand as a conceptual handle for a personal scale resource that does high performance computing. A good thing. But, from what I understand, I could open up shop tomorrow and sell Tesla Personal Supercomputers just by building something that conforms to something close to the minimum spec, and this is not a good thing. With no controls on the brand, it is inevitable that it will creep away from the standard. You’ll always have to ask the question, “yeah, but who’s Tesla Personal Supercomputer is that?” and the value as a shorthand for the concept disappears.”
NVIDIA got in touch with me to clarify this a bit, and it looks like there is more to certification than I gathered from my initial interviews with them. First, there is a badging process by which a company can earn the right to display the “Tesla Personal Supercomputer” badge. To get that badge a system must meet a set of hardware standards (minimum quad core CPU, at least 4 GB of system memory per GPU, and so on), and the system must pass a set of tests that are evaluated by NVIDIA, along with some other legal requirements.