Petz, Re: "Eliminating the huge D1D (175,000 s.f.) is probably going too far -- a fab this size is obviously being used for production as well as development."
D1D hasn't even finished construction, yet. Try driving by Evergreen Road in Hillsboro, OR, and you will see the construction. It doesn't look like it will be finished this year, either.
Did you know that? Probably not. Therefore, I would assume countless other errors and inconsistencies in Dan3's list as well, especially in his square footage numbers. Some of them simply seem too big, which leads me to believe they are square footage for the whole facility, not just the clean room.
I bet if you had a much clearer picture of Intel's facilities, you'd probably come to the same conclusion that they have: the market will recover and grow at xx% per year, and we need y more fabs in order to meet this demand in the coming years.
Why fight it? Why use Dan3 crafted data with enough holes such that three people have made corrections so far? AMD may be far more open with their data, but they also have a lot less data to be concerned about. Intel has so many fabs, and so many are currently undocumented in terms of their fab space or capacity, or even if they are still in operation, or if they have been shut down or converted into office space. Yet I've seen all kinds of guess work and speculation here made using the finest details about Intel's manufacturing capabilities, their costs, and their future ramp abilities. I'm wondering how much of this data is reliable, and how much is still current.
I doubt that Dan3 can even vouch for the reliability of his data. Much of it did come from a previous conversation that we had several months ago where I was able to find a lot of public information. However, there were also a lot of holes, and Dan3 seemed to fill these in with arbitrary numbers, or numbers that came from web pages that I couldn't ascertain as reliable sources.
Therefore, I appreciate your further analysis, but even with your eliminations, it's too hard to tell if you are correct with your assumptions (and most probably you are wrong). Sorry to say, but I doubt this conversation will be able to be settled without a simple difference of opinion in the end.
wbmw |
