Ex- and all, lead times were 7-8 months at worst, (best recollection) in the 1995/6 time frame, and then the qualification times at the fab were quite long (3 months, I think). The combined times got as low as 3 months during the 1998 doldrums, because AMAT had no backlog. (I?m talking here about PVD systems, I don?t know that much about other product lines.)
Right now, I suspect booking for ?slots? for systems is about 6 months out, and then there?s the 4 months to build, ship, and qualify. That?s best case, and might be as long as 14 months. Were you to call AMAT today and order a PVD system for ?quickest delivery? I bet you dimes to donuts you?ll have the system running in June ?01. Not before. (I'm guessing, I don't have an AMAT insider to get systems information from. But the very high bookings rate will surely cause the lead times to explode, because the backlog grows so dramatically!)
The labor market for tech folks is tapped out here in central Texas. To make matters worse, everybody?s hiring: Dell, AMD, Motorola, Samsung, the works. Last I knew, pretty much all these had referral bonus programs, some extending to ?anybody who refers somebody we hire? rather than the usual internal bonuses. College hire programs are seeing a dramatic reduction in acceptances, mainly because nearly every grad is getting several to many offers. In any case, college new hires don?t affect production for a while, so they?re not that useful to the bottom line right away.
I suspect that human resources are the true system throughput constraint at the moment.
Next would be buildings, then qualified suppliers. Although the inventory and logistics issues are very difficult! It only takes one unavailable part to delay a whole order.
On the fab throughput issue, lithography is usually the (designed-in) constraint, so adding tools elsewhere has a small to negligible effect. Until litho is maxed out, and then you add more litho tools (as you pointed out, these babies are hard to come by!), adding more equipment elsewhere is an expense that goes to the bottom line, often without improving revenues. Fabs continue to do it, I don?t know why. (I don?t mind, this ?effect? is responsible for a nice little consultancy.)
Mitch |
