RealMoney Silver - AMD: The More Things Change ...

AMD: The More Things Change ...
By Bob Faulkner
RealMoney Contributor
12/19/2007 3:20 PM EST

Sorry, but I must take issue with a video over on TheStreet.com indicating AMD (AMD) has adopted a new "strategy." It's the same strategy it has always been, but a different focal point.

AMD Reverting to Failed Strategy?

The authors suggested that because AMD talked about addressing the broader desktop and notebook markets at their recent analyst meeting, that this was a shift away from the high-end server space in which they had so much success in 2006. This is simply a matter of, "When all you have is a hammer, everything looks like a nail."

The reason AMD was able to garner market share and embarrass Intel (INTC) in 2006 was quite simple -- it had something Intel didn't. AMD had designed the first true dual-core x86 processor. That architectural approach boosted performance and reduced its thirst for electricity, and those are characteristics that everyone running an IT server farm finds extremely attractive.

Intel was trumped and admitted it by virtue of its lame attempt to put two processors in the same package and call it a dual-core solution. As I said at the time, that's like slapping antlers and a red nose on your dog and trying to pass him off as Rudolph.

AMD had succeeded because Intel was caught with their pants down, and that's no longer the case. Battered, bruised and thoroughly embarrassed, Intel headed back into its cave to fix itself, and it has done that quite well. The company's products at virtually all levels surpass what AMD has to offer from a price, performance and power-consumption basis.

Not only has Intel surpassed AMD with its parts designed for high-end servers, the AMD solution has a bug. Granted, that bug will be fixed and the product will start shipping in 2008, but it's still not available. Maybe more important, several of the Taiwanese motherboard companies are suggesting that Intel is holding back on its next-generation (quad-core 45nm) parts until AMD launches its delayed Barcelona parts that are quad-core but at 65nm. That immediately puts AMD at a cost or pricing disadvantage.

So why is AMD talking up addressing the broader mass market opportunity? Because it has to. AMD has always been in the desktop and notebook segments. If I'm an OEM, I want volume pricing and I'm not going to get it by limiting purchases to one piece of the puzzle.

But the more critical issue from AMD's prospective is they have two fabs -- Fab 36 and the forthcoming Fab 38. Those facilities come with multi-billion-dollar capital expenditures and are running about $300 million per quarter in depreciation. That fixed expense has to be spread over as many processor units as possible, and there simply aren't enough server sales to soak up that depreciation.

AMD has no choice but to be in all segments of the market. Were they not, they'd wind up losing many times more than what they have done recently.

The combination of not having a competitive opening (a la 2006) plus being at a competitive cost disadvantage with late parts on an older-generation process make this a very difficult time to be an AMD investor. AMD woke up the slumbering bear by hitting it between the eyes with a two-by-four. Now they're paying the price and will continue to pay because the bear is angry.

AMD's not going out of business. As I noted last week, their OEM customers won't let that happen because their viability makes Intel more attentive to the OEMs. But you really have to stretch the imagination to create an investment case for this stock.