Brian, Cary, threaders,
Time for the year-end predictions, eh? I've been doing a little math-modeling on the side concerning some of the technical fundamentals in IC manufacture, and hence IC equipment. I'll put some of these thoughts out here for talking purposes, but keep in mind, they're not assumption-free. And for the most part, I expect to be found incorrect.
It is now my opinion that IC manufacture as we know it has seen its Golden Age. The peak of $206 B revenues will never be seen again, in constant dollars. The rate at which technology continues to shrink geometries, current (momentary) barrier notwithstanding, is now running away from demand; and due to the foibles of compounding, I don't see demand ever catching up again. For Si based stuff, certainly.
So what will happen? Nanotechnology with broaden out, include bio apps, plastic circuitry, you name it. And Moore's Law will still hold, but in an ever more aggressive form, Kurzweil's Law of Accelerating Returns (a more general phenomenon encompassing Moore's Law as a special case).
Large equipment firms that can throw off cash starting now have the potential to reinvent themselves and catch the next wave. However, given the way bureaucratic capitalism works, I feel these firms will go the way of the dinosaurs, and the next mammals will appear; from where I don't know yet, or I'd be investing in them! (And although AMAT is one of my favorite firms, I have to place them in the former, not the latter, in the long run.)
Overall, global fab square footage will continue to shrink, even with new fabs announced. Some of the older fabs will retool to related nanotechnology, some not. But circuit-building will need fewer companies (ST Micro says 6, I think maybe even fewer big ones) and less space. Money for tools may grow, but even there I'm pessimistic in the long run. Today's tools have much longer lifetimes than they're currently used for, and they'll work for many nanotech items with minor retooling.
I also think we're in the early stages of a "new materials Renaissance" much like happened in the 40's to early 60's, which wave propeled us into the Si-based era we all now enjoy. These new materials will drive quantum computing, optocomputing, biocomputing and more.
I've got other predictions, but I'll start with this set...
Mitch |
