Wednesday, May 31, 2006

AMD and ATI: Place Your Bets Now

Forbes.com has reported that Apjit Walia, an analyst with RBC Capital Markets, thinks that AMD might be considering a merger with ATI. There are a lot of qualifiers in the story, but it was a slow business news day, so it got a lot of play. Given that the source of the story is a single analyst who doesn’t cite any substantive evidence that a deal is seriously being considered, there’s not a whole lot of “there” there.

I don’t believe that this merger is going to happen any time soon, and the reason has nothing to do with graphics. NVIDIA is by far the largest third-party supplier of motherboard chipsets for AMD’s processors. I don’t know this for sure, but I suspect that they sell more chipsets for AMD processors than AMD does. NVIDIA’s chipsets have been instrumental in making AMD’s processors competitive with Intel’s, because raw processor speed is worthless if you can’t make it work in real-world PCs and servers. The two companies have been critically important to each other’s success: AMD’s processors made NVIDIA a serious player in the chipset business, and NVIDIA’s chipsets helped establish AMD’s processors as the best choice for hard-core gaming and graphics applications.

If AMD and ATI were to merge, NVIDIA would almost certainly pare back its commitment to AMD. That would be especially problematic for AMD now that Intel is ramping up to deliver its Conroe and Merom processors later this year, which will stack up very well against AMD’s models.

For this reason, as well as for the reason that acquiring ATI really doesn’t buy AMD anything, I don’t believe that an AMD-ATI merger is in the cards. But, I’ve been wrong before, so never say never. If a merger or acquisition should occur, however, the NVIDIA chipset problem won’t go away. In my opinion, an AMD-ATI deal makes less than no sense, but companies sometimes make emotional decisions and then rationalize them after the fact. Stay tuned.

No comments: