>> I would suggest that to
the end user the greater density and more functions on a devise offer greater net
benefit/price than the previous generation. To simply produce data that suggests
that ASP's normally trend upward is misleading. To draw the conclusion from the
data that chips are not commodities will result in a poor investment decision.<<
Then we are trying to make two different points. My point is that, as ASPs increase, chip makers have more money to spend on fabs and AMAT makes more money. As ASPs decrease, chip makers have less money to spend on fabs and AMAT makes less money. For purposes of this discussion, I don't much care what happens to the chip customers.
Yes, as it happens, the chip customers *do* get more functionality for their money. That's one reason why they're willing to continue to buy new chips, and why chips therefore are *not* commodities in the same way that, say, wheat is. If Moore's Law ever breaks down, we'll see a *huge* change in the industry's business model, but I don't see any signs of that happening right now, and Moore's law effects are not what I'm talking about here.
>>Finally, given Moore's second law, the profit structure to the manufactures of
the newest generation product is not linear and thus to suggest that manufacturers
reap huge profit margins as ASP trend upwards is also misleading.<<
That's an oversimplification. ASPs depend, among other things, on supply and demand. Can we at least agree that a supplier whose product is in short supply will make more money than one whose product is readily available? ASPs don't *cause* large profit margins, per se, but they are a metric indicating the supply/demand state of the chip industry.
Katherine |
