Dan, about your ruminations on Intel cutting CuMine prices to the bone, like 700's for $200, I think thats the direction that Intel will move, while simulataneously increasing prices on the lower-end CPU's, but it won't happen overnight.
Since Intel introduced the celeron, the pricing structure of the CPU industry has been distorted. This is because Intel had no real lasting competition at the high end. Their marketing strategy to brand the Pentium III as being far superior to the Celeron never really worked. Any fool knew that a PIII-500 is only worth 20% more than a Celeron, but if they wanted something faster than a Celeron 500 they HAD to pay up. The ratio of top price to bottom price grew and grew only because there was no competition in the high end.
Obviously, that has begun to change. I say "begun" not because AMD is not competitive performance-wise but because it is not fully competitive production-wise. Also, AMD is not yet competitive at the extreme high end, server chips.
I think AMD will have an impossible time selling more than 2M Athlon chips per quarter at an ASP above $250, even if they are all 750's. (Hey, I'm not complaining, thats half a billion in CPU revenue just from Athlons!) Intel will drop prices to meet the competition, AMD. And as AMD starts making fewer K6-2's and/or K6-2P's, Intel will start raising prices at the low end. Basically, in a truly competitive environment, the cost of CPU's will more closely approximate their utility to the consumer. Why, because basic free-market economics works that way. Of course, if either AMD or Intel manages to achieve achieve a monopoly at the high end, those CPU's will sell at inflated prices.
I expect the price ratio between the top and bottom ends of the CPU spectrum to compress from 10:1 to about 3:1. The result for Intel's profits will be disastrous; in the long run, if AMD also manages to create competition in the server arena, Intel's pre-tax profit margin will drop to 20%.
Petz |
