Elmer Re... There was more demand than Intel could supply in Q4 .<<<<<<<
Don't you see the problem in that? If that demand resulted in Intel upping it's cpu total by 10 - 20% fine. But the fact that Intel just sold slightly over 32.5 mil. cpu's, and had to run 5 - 7 fabs full out to do it, does equate with yield problems; or the die are too big to allow the fabs to meet demand adequately. Every manufacturer ideally should have a 20% production capacity to meet potential increased demand, or overcome downtime. The fact that Intel couldn't meet demand, during a downturn portends looming trouble. Now I realize Intel is in the process of building 3 superfabs, and won't have problems meeting quotas, but the need for so much fab space, for only for 4 times the volume of chips compared to AMD, spells trouble in a price war, and isn't something one should be bragging about.
still AMD couldn't sell out much more than half Fab30s capacity.<<<<<
How does that possibly equate to yield problems? AMD produced both TBird and XP models at Dresden. 5 - 6 mil units at 50% fab capacity. Intel used 2 fabs to produce
8 mil P4. On the surface, AMD's yields look a lot better. And that is the problem with your claims. They don't seem logical, and your only answer is to say, P4 is a big die, so it should have poorer yields/wafer. That just proves Intel's process is better. Huh. I would say that it doesn't necessarily mean Intel's process, or percentage of good die, are poor; but it does mean, in the total, whether it is the big die or poor yields, something is causing Intels total yield/wafer to be poorer than AMD's. You say "AMD still couldn't sell out more than half" as if AMD had a big downturn in sales, when in fact, AMD sold a record amount of chips in the 4 q-01. It is just that their capacity, or yields increased faster than sales, which can be a good thing. |
