Replies to post #23020 on Advanced Micro Devices Inc (AMD)

[yourbankruptcy]

Intel definitely will have overcapacity. As we read today, IDC initially expected Itanium server market to be $30 billion in 2004. Intel had build the capacity for 30 bln market. Now IDC decreased this number from 30 to less than 8 $bln. What to do?

[kpf]

Thats all...


Sure. Baking chips is like baking cakes. An oven is an oven, flour is flour and sugar is sugar. A fab is a fab, silicon is silicon and copper is copper. Impressive you found all that out in no time. Keeping up this pace you will have what it takes to start your own fabless CPU manufacturing empire well before 2005. Make sure that none of these eggheaded phycicists will ever cross your way and post again for IPO-data.

K.

[Mysef]

Tiger64,
re:capacity

I havent read the article but...

Are you assuming all of this fab capacity is for CPUs?

Only a few companies with fabs are successful in producing CPUs.
I'm guessing the article is talking about the flash glut that may or may not arrive in two years?

Mysef

[CombJelly]

"Why not go joint venture with some of these plants to expand capacity"

Several reasons. Remember when AMD tried to fab at UMC? Now there are all sort of rumors as to what the problems were, but I will point out that the fastest complex device that UMC or TSMC (can't remember who fabs for Via offhand) produces is a Via processor at 1.4GHz.

About the only companies that possibly could fab a bleeding edge processor for AMD is Intel, IBM and possibly Motorola. Motorola is trying to dump their fabs, no real long term possibilities there. IBM might be able to do it, but they might not. As far as Intel...

For better or worse, it takes a process tuned for high speed and tight intergration between the process and the design to get a modern, high speed processor. When Intel chose to push the performance envelope with the Pentium Pro and its descendants, the handwriting was on the wall for fabless processor companies. That is why they died.