"I imagine that's how AMD can price its chips below intel's and still have a profit"
Nope. AMD has lower costs over all, and this has been true for a long time. It has far fewer people, a lot fewer buildings, less corporate structure. If processor ASPs are at $100 for AMD it can make good money. If Intel's ASP ever drops to $100, they will bleed deep, wide rivers.
"Is there a significant cost improvement just from 300mm vs 200mm even at the same 90nm"
It depends on what you mean by "significant". It costs more or less the same to process a 300mm wafer as it does a 200mm. Not precisely, but close. The wafers cost more, but out of ~$3k, a wafer might be as much as $300, maybe less. However, you get 2.25 t0 2.5 more raw die per wafer.
So your raw die cost might drop from $10 to $4.
Edit: ok, make that $15 to $7. Assuming 150mm^2 die and a processed wafer cost of $3000.
Re: AMD can make 50% more dual core CPU's on 12" wafers than it can make single core on 8" wafers.
Huh? 300mm wafer is 225% larger than 200mm. The number of chips goes up 240% but I am not sure how that is calculated.
The wafer area goes up by the square of the diameter, but the unused wastage near the circumference goes up only linearly with the diameter. I think Petz is implying that a dual-core chip is itself 160% bigger than a single-core.
>> Nope. AMD has lower costs over all, and this has been true for a long time. It has far fewer people, a lot fewer buildings, less corporate structure.
Thanks for clearing that up for me. I imagine that lean operation comes from surviving the many years of low profits. I don't suppose there is a point in asking why that same lean corporate structure didn't help the flash memory operation.
Thanks and good luck to all who responded to my posts. I can now appreciate the source of enthusiasm for the stock price.