InvestorsHub Logo
icon url

kpf

08/19/03 3:19 PM

#11452 RE: subzero #11430

sz - Do the math

Although 30-35% were certainly accurate for some period last year, I doubt it is still accurate today:

1. I encourage you to look thoroughly at the work-in-process numbers of the last five quarters. Plenty of room to play with inventories by means of this position which cannot be easily understood using just easy math. (see below)

2. We have already seen models specified with lower Vcc for Thoroughbred, Bartons and K8, indicating less aggressive processes tending to yield significantly better than the aggressive process AMD had to use before to get at least something competitive out of the fab.

1+2: The output from 2 goes into finished products only after packaging, valued at market prices - if you keep the raw die it is in the work-in-process figure, valued at its cost.

But even taking all this into consideration, your figures are way closer than what most people think of yields.

K.





icon url

Andy Grave

08/19/03 4:27 PM

#11461 RE: subzero #11430

The 30 - 35% yield falls out every time

I guess you believe AMD officials are foolish enough to give out all the necessary information such that a clever person like you can easily calculate yields. That's a possibility. But...it's also possible AMD officials are not so foolish and you are not so clever. You may want to consider that possibility also.

Andy Grave