Advanced Micro Devices - Moderated (AMD) Message Board

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.

Technology Stocks : Advanced Micro Devices - Moderated (AMD)

AMD 203.14

-0.8%

3:59 PM EST

Public Reply Prvt Reply Mark as Last Read File

Previous 10 Next 10 Previous Next

To: Monica Detwiler who wrote (79379)	5/5/2002 12:30:00 PM
From: pgerassi	Read Replies (2) of 275872

Monica: Intel sells fantasy chips. It's part of the wait till next period (you pick the period) answer they always lately seem to give. Well we wait and wait and they still have not made it. When can I get a real 2.4GHz P4 in performance comparative to a P3 running at 2.4GHz? Not some Intel "spinning its wheels" CPU that maybe gets 1.6GHz on real performance on most software, if that. In the whole world, not some restricted tiny subdomain. Here is where that scaling factor works. It takes the speed of the CPU and runs it through a simple formula to estimate its true performance running on a type of software. The former means nothing wrt to performance except that higher numbers of the same exact CPU mean somewhat that the higher number has more performance than a lower number. It does not compare to a clock cycle on different CPUs at all. P4 has a very low gear and AXP a high one. A revolution by P4 goes much less far than a revolution by AMD. Like 3000rpm in a muscle car is much faster than 4000rpm in a go kart. So you need scaling as a first order approximation to relate a given CPU's clock to performance. And the supplied scaling factors show AXP is 2.4 times as powerful as P4 per clock cycle. As to model numbers, they give performance of the target CPU compared to AT LEAST model number of megahertz of a Tbird giving the same performance overall. The "at least" throws you and many others for a loop. You and many others assume "equal to" rather than "greater than" and no where is that stated or implied. So a Athlon of model 4000 (about a 2.5GHz AXP64) would outrun a Tbird running at 4000MHz or 4GHz which outruns a P3 running at 4GHz which easily outruns a P4 of much higher than 4GHz. Given the scaling factors, P4 would need to go to 6GHz at a minimum to begin to match. Yes some benches would make it faster, but on most software it will be much slower. Of course all of this would be unnecessary if Intel increased the IPC rather than substantially decreasing it for P4 over P3 like all their generation changes before. Selling 4 times the CPUs for Intel is a step down, they used to sell 9 to 1, not too long ago. And that ratio is lowering as we speak. Pete

Report TOU Violation

Share This Post

Public Reply Prvt Reply Mark as Last Read File

Previous 10 Next 10 Previous Next