I am still of the mind that there would not be much difference in parts cost between a 50% utilized fab ramping up, and a 50% utilized fab that is mature.
Say Hertz has 1,000,000 cars with 50% rented out and 50% parked in their lots. Meanwhile Avis has 500,000 cars with every one rented out and just ordered 25,000 more. The cost of Hertz's 500,000 on the road is the cost of their 1,000,000 cars. Avis's cost is lower.
the ramp-up would have lower yields, higher engineering cost (for the lines being validated, qualified). There would still be equipment that is purchased but not producing for revenue, etc....
It isn't clear to me that the ramp-up would have lower yields, running the same process on the same equipment. And didn't you postulate a "50% utilized fab that is mature"? "There would still be equipment that is purchased but not producing for revenue," a lot of it, 50% in fact.
How can one tell whether this notion is accurate, or just another hare-brained idea ? Would the answer be in AMD's q2 and q3 reports ?
You'd have to look hard to find the "50% utilized fab that is mature" in AMD's reports. I don't even think Intel has one. But if you want to compare costs, look at gross margin. AMD's is already about as good as one could reasonably expect.
My guess is the benefit from fab 36 will not be apparent in q2 and q3.
It should be apparent in YoY revenue growth. |
