re: 7.5B capex in 2001:
Intel has reiterated this number about a dozen times recently. Can they really do it?
1. Intel is now a conglomerate, in a variety of markets. Supposedly, the advantage of that, is that weakness in one area should be balanced with strength in other areas. But, Intel has just said, and overwhelming evidence from other companies also says, that all their markets have demand much weaker than expected.
2. Intel is looking for ways to cut expenses. They recently said they would be cutting "discretionary spending", but this will only yield a few hundred million. Ultimately, all spending is discretionary. At some point, they will need to look at the capex budget if they are serious about cost cutting.
3. Intel is betting on a quick V-shaped economic downturn, so that 2001's capex spending will create maximum production (at maximum efficiency) in a market where demand has rebounded sharply. I remember back in 1998, when several semis had idle state-of-the-art fabs. That is a very expensive mistake to make. If we get a U-shaped downturn, Intel should be cutting capex.
4. On the other hand, we are in the midst of several crucial changes in fab technolgy (300mm, copper), and this may keep capex high. In the 1998 downturn, the only tech change chip companies made was to continue shrinks. They put 300mm and copper on hold. Today, that is harder to do. |
