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 : Advanced Micro Devices - Moderated (AMD) -- Ignore unavailable to you. Want to Upgrade?

To: that_crazy_doug who wrote (12127)	10/9/2000 2:50:01 AM
From: Charles R	Respond to of 275872

<The only real problem it had was that it was lower speed, and lower performance at the same speed? What do you consider real problems? Not as pretty a color? Choking hazzard if swallowed? > I do not consider lower speed to be a problem if the lower speed falls within shippable MHz window - as it did for K6(it, however, is a problem to AMD ASPs, of course) I consider significant reduction in performance as a problem(something in the 20-30% range or greater). I also consider incompatibilities that manifest to an extent that an OEM does not ship as a problem. If a certain system builders is displeased with the color or swallows it and chokes on it I may consider it a problem - for the system builders ;-) <It's a processor, it's primary purpose is to process, it didn't do it at as high mhz or as effectively at the same mhz, those are the two biggest problems it could have besides just doing it entirely incorrectly. > Duh! Lack of MHz and somewhat reduced performance was a big part of the reason why K6s were sold at a discount! <(There were also more driver compatibility issues with it, regardless of whether you're willing to believe it). > At no point did I say that I did not believe there were some compatibility problems. Me thinks it is pointless debating with a person who does not understand the subtlity between having a compatibility problem with some odd stuff that does not matter to most people and compatibility problem that prevents a large OEM from shipping systems. <It's funny how you refuse to break anything into details, ..> My dear sir, I post beaucoup details about pretty much any question that people ask. Might I suggest I cannot be held responsible for me not knowing what level of detail I need to go to get some posters to understand the point. <..you say I'm putting words in your mouth ..> That you do and that is not a good habit. <...or that you disagree, but you never clarify what it is you meant, ...> I think I do unless I am turned off by the tone of the post. <...or what part I misinterpreted, so what part do you disagree with: 1) That the major parts of performance are mhz, ipc, and software optimization? > This is pretty elementary stuff. No disagreements on this one. <... or 2) That you don't know how high mhz the p4 will go to, ..> This was not the point. I do say that I do not know how far P4 will scale. But that is not an issue that was under contention. The point I was making was that, at launch of P4, Intel will not have performance leadership. This has nothing to do with the frequency scalability which will not be an issue until sometime mid-to-late 2001. <...and that you don't know how well the software will benchmark out. > I am aware of how the software is benchmarking out now (it is not looking good for P4). The point I was making was that software improvements which may come along in the future may help P4s cause. I do not see much of a chance that that will happen at the release and hence my confidence level about P4 not looking good coming out of the gates. If you take it mean I don't know how well software will benchmark out that is fine but just make sure you say it in the proper context. My comments about P4 underperforming are strictly related to the Q4 launch. I expect a lot of things to get better for P4 starting late-Q1 and early Q2. The absolute MHz delta is likely to be too large to ignore starting that time frame. I have often said that P4 on 0.13 will send Athlon family to mid-range unless P4 hits a scaling problem (which I think is a finite possibility).