Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Politics : Formerly About Advanced Micro Devices -- Ignore unavailable to you. Want to Upgrade?

To: Bill Jackson who wrote (43753)	12/21/1998 7:40:00 PM
From: nihil	Read Replies (1) \| Respond to of 1574005

re: absolute limit I think you are right. A normal curve is a continuous approximation and limiting distribution for the discrete bi(poly)nomial distribution that characterizes a process with a finite number of discrete elements. As complex as a cpu may be, I guess it actually is finite and even if defect free its maximum rate of operation would depend on how much heat it could handle with tolerable MTBF. Thus no 10GHz chips. Of course, the inherent and unavoidable variability in dice from any single wafer, however small, may be enough to make it common for different speed grades to occur on a single wafer. I have been told that Intel does not serialize and index chips, and perhaps this is a reason. If demand for a 450 MHz chips is less elastic than demand for 400 MHz chips (as seems likely) it could easily be more profitable for Intel to downgrade surplus 450MHz chips to 400MHz, rather than reduce the price difference. If there is no way for a lawyer to recover individual wafer speed distributions and speed grades, I think he will not be able to prove his case. Unless these data are recorded and discoverable, your lawyer friend will not be able to prove that Intel systematically discriminates in price on identical products. Intel can legally charge different prices for different (different speed) chips. To prove that a 400 MHz chip is not different from a chip based on a die of the same speed from the same wafer is impossible, especially if Intel has jiggered the 400MHz chip so it can't be overclocked (i.e. the slower-clocked chip is functionally slower than the faster-clocked chip, regardless of the hypothetical potential of the two dice). To prove that Intel is engaged in an illegal spoofing of chip differences will be very hard, especially given that Intel frequently changes the relative prices and the price level of the two chips.