Is there some other metric or data you can point to that indicates AMD has a better production process than INTC?
Me...
As everyone should know by now I'm not technical so basically something has to hit me over the head for me to pay attention to it. As such I basically rely on empirical judgements to guide me. When I look at the number of fabs INTC has and try to reconcile that with the shortages (yes I know INTC says chipsets are produced on older processes) INTC is claiming I can only come to the conclusion that something isn't right with INTC's production processes. Besides the shortages in chipsets I seem to remember Andy saying in the MidQ that INTC was running at full capacity and had less inventory than they wanted. Again, another red flag.
On the other side of things it seems pretty apparent that since the "E" stepping AMD has enjoyed much better bin splits and yields. I'm saying this primarily because of the lack of supply problems. Again, I am aware of the latest Asian supply problems, but since no one else is having these problems they're probably localized. From my point of view AMD is having much less of a problem supplying markets than I had expected. And no it's not because of a lack of demand. This at a time when INTC is having admitted supply problems which should be adding increased pressure onto AMD's ability to supply.
Given the design wins AMD has garnered and the large increases in server sales, without a big improvement in AMD's production processes(other than just 90nm) I see no way AMD could be performing like it is. And this doesn't take into consideration the additional demands Turions have been placing on production, nor does it explain why AMD hasn't been forced to pare back Sempron sales given the restructured priority picture.
All in all I've concluded that SOI and DSL are worth a lot more than INTC claims. Further, I now suspect that since the "E" stepping changes the bin splits have improved to the point where AMD chips are so far ahead of the "sweet spot" that AMD is able to sell virtually every chip it produces.
The way I look at it AMD now has production processes in place that allow it to raise the "sweet spot" at will which should keep AMD ahead of INTC at least until 65nm chips start to appear. But given INTC's past history of problems with 90nm that's far from a given.
There is a limit to how many chips AMD can produce even with the improved process methods, and I expect we'll start to hear more about that as we get further into q4. Right now I'm thinking AMD probably produced about 10 to 10.5 million chips in q3, pretty close to absolute max production of about 11.5m chips which they should hit in q4.
So you see that while I don't have the technical background I do have a method behind my madness. I'll leave the technical stuff to you guys.
The real problem I have with the technical stuff is that it's been my experience that it's too easy to wrap up problems in the mumbo jumbo and obfuscate the obvious. I've found that, as a rule, I'm much better off using my eyes and my two bits of gray matter when it comes to figuring out what is really going on.
