AMD now making GPUs and semi-custom chips at GlobalFoundries

Speaking at a Credit Suisse Technology Conference, AMD CFO Devinder Kumar noted that this year AMD is making its first GPUs and semi-custom chips at GlobalFoundries. Previously, AMD relied on TSMC for the production of GPUs but it seems this relationship is coming to an end. Devinder doesn't really go into the details, but he mentions that while the bulk of AMD's products are 28nm right now, they'll have certain 20nm products in the near-future before moving to GlobalFoundries' 14nm FinFET process.

John Pitzer-Credit Suisse Securities - Credit Suisse Securities
Can you talk a little bit about kind of your roadmap around Moore's Law; a lot of discussion at the industry level about FinFET, about whether or not Moore's Law is slowing. AMD has been a little bit more absent in the conversation. That might just be a function of the embedded business, not necessarily needing to push Moore's Law as aggressively as perhaps graphics and the CPU business, but what do you think about your roadmap of getting the FinFET and moving down Moore's Law and the expense of doing so?

Devinder Kumar - SVP and CFO
AMD has been absent because we are a fabless company. And typically fabless companies are not the ones that have to worry about the process technologies. I think that's for the GlobalFoundries to go ahead and figure out. And GlobalFoundries did something very interesting, the alliance with Samsung, I think that's going to help both Samsung and GlobalFoundries and us from an overall standpoint because it makes FinFET available to a broader set of customers in the industry as opposed we have one or 2 players that go ahead and have the technology.

From our standpoint, given our product portfolio and the customers and the market that they're pursuing, we do not need to be at the bleeding edge of technology. We need to be at the leading edge, we don't need to be at bleeding edge like we used to when we had the PADs, we pushed the process technology equation. In the compute space, we have shown the ability from a power, efficiency and performance standpoint and architecture and software investments to have competitive products and that is in some sense less costly if we pick and put it that way than investing millions of dollars in technology from an overall standpoint.

So from an overall standpoint, today the bulk of our products are 28 nanometers. We will have certain products in 20 nanometers and then we'll go to FinFET from there, but our partnership with GlobalFoundries is where it comes into play in terms of what's the right point to go and intersect products with technology and introduce the parts out in the market.

You can read the full transcript at Seeking Alpha.