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: wanna_bmw who wrote (50541)	8/9/2001 5:39:17 PM
From: Jim McMannis	Read Replies (2) \| Respond to of 275872

RE:"many past micro-architectures have needed software optimizations to truly break away from the older designs" It remains to be seen if the P4s micro architecture is truely "break away" or broken off. Appears to me that the P4 was designed with two things in mind. 1. High clockspeed. 2. Rambus... no. 1 was a good marketing tool. MHz sells... no. 2 the second was a bad try at controlling the market. backfired but there is always backfill to SDRAM. Intel is doing there best to hide the weaknesses of the P4...AMD could have never done that good. Jim

To: wanna_bmw who wrote (50541)	8/9/2001 7:43:53 PM
From: jcholewa	Respond to of 275872

> If you think that, then you ought to extend your statement > to many of Intel's previous generations, because many past > micro-architectures have needed software optimizations to > truly break away from the older designs. I think the point is that the Netburst is the first Intel processor core to have significantly lower per-clock performance to the previous generation before optimization, and more importantly it is the first Intel processor core to not be the fastest processor at its release. The P5 was much faster than the 486 on a per-clock basis. Guesstimates had the 75MHz P5 at least equalling the performance of a 133MHz clock quadrupled 486 (well, "5x86"). So it was faster per clock that Intel's fastest 80486DX chip, and it was faster than the Cyrix and AMD 486 chips that met it at release. The Pentium II generally outperformed the Pentium MMX in per-clock performance, especially in floating-point stuff and rol (small thing) stuff. When the Pentium II came out, it was no slower than the AMD K6 or the Cyrix 6x86MX, its immediate competition. It had an equal or greater performance across the board and was <em>incredibly</em> faster in a nontrivial amount of benchmarks. (I ignored Pentium Pro here, but it is another excellent example) The Pentium III (though not an actual generation change) also had the distinction of matching its predecessor (PII) in terms of per-clock performance and it was overall generally faster than competing K6-2 and K6-III processors. Before the Pentium 4, there was only one instance where a new Intel processor class both underperformed the competition while being significantly lower in per-clock performance. This part was the original "cacheless" Celeron (which had actually a 32KB L1 cache but no L2 cache and no motherboard cache), Covington. This processor was pretty much a bust, and was reviled by the media all over, with about two places recommending it, and then only for hardcore overclockers (Tom's Hardware Guide was one of the two places). Intel had plans to eventually replace it with a version of the Celeron with L2 cache (the "Mendocino"), but the backlash prompted them to hasten this transition. The Pentium 4, at introduction, was slower than the Pentium III from a per-clock perspective in the majority of applications before optimization. Even with optimization, the amount of code that can be converted over is limited in quantity, and the Pentium III likely keeps a per clock advantage in most situations. Additionally, the competition had a higher performing part at the time of the Pentium 4's unveiling before optimizations are taken into account (and, as with the Pentium III, the competing chip was likely faster in the majority of applications and programs even with optimization taken into account). This is what we mean when we talk about the crutches on which this processor relies. The Pentium 4 is unlike any other high end Intel processor in history, in both good ways and bad. This happens to be one of the bad. -JC