You have good points, but they just amplify what we all know, the chip business is a risky hire wire act with huge entry barriers. Just the same, if the demand is there, someone will step up to the plate and do what is necessary to supply it. the reports say that existing capacity, even after the last years' robust buying of new equipment for .18 shrinks, is strained to the limit for many device types. Micros for sure, flash for sure, and others. there is a healthy debate on whether DRAMS will be strained mid-year, so jury's out there.
Intel has upped capital spending to $5B, up 40% from last forecast not too long ago. Atmel new fab in Dallas; Taiwan too numerous to mention. And, guess what, foundries exhausted, if you don't have a commitment now you are SOL. Has many companies rethinking their foundry strategy, where you lose direct control of your own destiny.
Now, demand. I say it can only increase and maybe exponentially. Everywhere you look chip consumption is meteoric, and thats just what we see today. It seems now every new application leads to 5 more. And over the next 5 years , beginning to increase later this year, the move to replace conv. TV's with HDTV, the only standard in 2006. These sets are chip hogs; think of worldwide TV purchases as every existing set becomes unusable.
In a market where demand is high, you build fabs to stay in the game or you're competitors will. Maintain or increase market share or die. |
