Well, yes, Cary; and that's one of the flaws in my model. Remember: "Models are always wrong; however, they are sometimes useful."
Here's what's controlling demand in at least some ways. The computers I have now are all at least 3 years old. They still do a fine job, have plenty of memory, and seldom crash and burn. I don't foresee replacing them for at least 2-3 years yet, which means I'm falling way behind the curve on current technology.
I don't see that as a problem, not like I would have back when a 4164 chip was the hottest memory around. Or even when the PIII chip showed up. I felt I needed the hottest stuff, because I could see uses for the new functionality and enlarged capacity and speed. Now I've got enough.
I'm out of the demand pile, for a while; and it's not about the cost! Heck, things are cheaper than ever before.
In any case, computers are only a small part of Si-based circuitry in use, and contemplated. Even if demand grows in that sector by 50% it can't possibly drive factory utilization up greatly.
I still hope I'm incorrect in some of this stuff, otherwise I'm the biggest pessimist around. (I don't like that.) But I can't be too far off, or we wouldn't see so much rationalization of capacity and stretching of capital equipment buying cycles.
Mitch (Still peering out of the bunker) |
