wbmw,
The same might be said for the price that AMD is paying to be "bleeding edge" with SOI in 2002.
Definitely. There is definitely a cost to being on the bleeding edge in Copper and SOI on AMD side and on 300mm on Intel side. But I don't have the data on which I could make a comparison on who made the right decision.
All I was saying that the decreased cap. ex. costs may be all due to lower prices on the same amount of equipment that Intel purchased in 2001, as the equipment is entering mainstream.
You have to ask yourself what competitive advantages these technologies offer the companies that are pursuing them, first, before you assume that they are paying too high a price for the privilege of being early to the party.
Good point. Let me take a stab at it. SOI is mainly for performance, 300 mm is mainly for capacity.
Some would say that 300mm offers lower cost. I think they are wrong. If you buy your 300 mm equipment in 2003 at discounted prices, you achieve lower costs, but if you buy your equipment for top dollar in 2001, you are not achieving any savings. From reading the CC notes, Intel does not seem to be guiding to lower costs due to 300mm.
The primary motivation for rush to 300mm is that Intel ended up with a processor which takes twice the die space of the previous generation, needs to ramp it up fast, at the time 200mm equipment is at the tail end of it's life for volume manufacturing.
The motivation for going with SOI (= performance) is obvious, since that's where the $$$s are, and for a small company such as AMD, it makes a big difference if they supply the bottom 20%, mid-range 20% or 20% of high end market.
Joe |
