You can't just add more microprocessors to the marketplace and expect the prices to remain constant.
Assuming, of course, the market does not increase.
People on this forum and elsewhere have raised the point that Fab 25 will be converted to flash, that eventually the Intel fabs will be retired, and that overall microprocessor demand is steadily rising. All of these may come true, but to me the important point is that they all must come true for AMD to be as highly profitable as so many hope.
I think you raise some very good points. For some time I held similar views and believed that there would be a significant commoditization of this market, but it hasn't happened.
After Intel's segmentation hat trick with the celeron (BTW, a brilliant stroke against AMD last year and great marketing)and possible further segmentation with the Itanic....I have changed my mind.
Intel and AMD still have too much invested to lose it by commoditizing the price for chips. Although Intel is remaking itself and diversifying, its core business remains semiconductor manufacturing.
I also am not of the opinion that thin clients or Internet appliances will significantly supplant the pc in either the consumer or business marketplace. They are great ideas, but will never replace the flexibility and multi-function capabilities of the pc. I also believe that this market will only increase.
Segmentation will allow Intel and AMD to supply chips for lower-end applications while delivering high-end and expensive chips to those who will continue to demand them. By providing performance distinctions priced accordingly, purchasers will accept these differences.
Who buys high-performance systems? At the home, not just gamers, who do often buy the fastest systems available. Many people will buy the fastest system possible in order to mitigate the expected obsolescence of their purchase. Also, a certain class of consumers will buy the fastest and more expensive because they believe there is value in the additional speed. Higher-end systems are also often configured with additional peripherals, bigger monitors, scanners, etc. This same marketing philosophy applies to automobiles where higher-end cars come standard with amenities which are options at the lower-end.
Businesses have never had a problem spending significantly more for computers than home consumers. In a more technical world, we need better and faster computers. No, my secretary doesn't need a faster computer to type a memo, but considering all the applications she is running...she needs a fast processor with lots of memory and a fast harddrive. Add to this the bloatware effect of MS software releases....and a new computer is in order every few years; I know our IS department always buys the fastest possible to also mitigate obsolescence.
AMD and Intel will have lower-end processors to reach that market as well. But consider how much Intel is getting for Xeon processors, and that the 800mhz Athlon is about $800, while the k6-2 500 is well under $100.....seems to me that this segmentation strategy is working very well.
Yes, the demand for processors must increase, but our entire economic system is based upon the expansion of markets....right? Companies are evaluated on the basis of expansion and growth.....so I don't have a problem with assumptions of growth in the semiconductor market.
-Scot |
