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.

Technology Stocks : Intel Corporation (INTC) -- Ignore unavailable to you. Want to Upgrade?

To: Joe NYC who wrote (150868)	12/3/2001 6:09:01 PM
From: wanna_bmw	Read Replies (1) \| Respond to of 186894

Joe, Re: "I think the lack of specs on how an instruction makes it through the decoder, and how the micro-ops are generated may be the Achilles' heel of this project." I don't get what you hope to prove? Do you want to pretend you're all CPU engineers so that you can show Intel how to design a better CPU? I don't doubt that the Pentium 4 has many Achilles' heels, but what does it matter of 99.999% of the applications out there don't hit those corner cases. Part of the idea around optimizing a CPU is to aim for the largest and most popular kinds of applications that need the performance the most. If you are able to get a 10% performance boost in 99% of code by making a change, but the penalty is 20% in 1% of the code, would you make the change? I'm thinking that Intel engineers probably did, and on several occasions. If you take the aggregate sum of the whole, you can see that when software is favorable to the Pentium 4 (like in Lightwave 7b), it can actually operate clock for clock better than the Athlon, and still be able to clock much higher. On most applications, it takes the extra frequency to make up the difference, and in a few very rare programs, even the frequency doesn't make up a thing. I think a design like the Pentium 4 only works if there is enough of the first kind of application to make it compelling. Right now, there is not, but you can bet that there will be. Exposing the Achilles' Heels of the Pentium 4 does nothing more than prove the point that the chip has weaknesses, and that's not a big deal if you can work around them. So again, I'm trying to find out the point of all this. Are you simply curious at finding out what the biggest pitfalls are (in which case, you can probably pick up a Pentium 4 spec, and read most of them there), or are you looking for more ways to raise a fuss. I think Kapkan's plan was always the latter. And he'll make his fuss, and the rest of us will ignore him. wbmw