>>Can you give me a listing of any new announced fabs that weren't
planned/announced prior to 1998. I have not heard of one newly
announced fab expansion so far this year, but I have read/heard about
numerous fab delays and fab closures due to the overcapacity issues
that are still prevalent and will be into 2000. <<
I'm not aware of any newly announced fabs, but that doesn't mean they don't exist. That isn't what I said, anyway. I was simply challenging your claim that there's only one fab under construction in the world right now.
>> All of the delayed
fabs that were previously announced were suppose to be 300mm, and
it seems the reasoning given for the delays was due to current
production overcapacity and an incomplete 300mm equipment set that
is unproven. <<
No argument there. Also, it appears that, thanks to shrinks, the economics for 300mm aren't compelling enough yet:
news.semiconductoronline.com
>>As for INTELs' capital equipment expenditures for '98
I haven't heard of any huge single orders from one specific vendor,
which leads me to believe it is being used to selectively upgrade
outdated fabs/equipment to 0.25uM capability and possibly for R&D
evaluation purchases of newer technologies and 300mm equipment. If
you information to the contrary please elaborate.<<
As Ian pointed out, Intel vendors are not permitted to make announcements. However, Intel has maintained a fairly consistent construction schedule for several years. They usually have three fabs under construction at any given time--one almost done, one in the early stages, and one in between. Their current level of capital spending suggests that they are maintaining this schedule. IMO, since they can afford to spend without exceeding their historic spending-to-revenue ratio, and many of their competitors cannot, Intel would be out of their collective minds to pull back any further.
Given that a complete fab costs $1.5-$2 billion, $4.5 billion will buy an awful lot of evaluation tools and upgrades.
Katherine |
