SI
SI
discoversearch

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)
INTC 36.15-0.6%Dec 24 12:59 PM EST

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: Raymond Thomas who started this subject8/19/2002 10:51:21 PM
From: wanna_bmw  Read Replies (3) of 186894
 
2.49GHz overclocked Athlon against Pentium 4 2.53GHz?

hardocp.com

The scores show an interesting picture. The idea of the article was to test the new ATI Radeon 9700 Pro, but they were able to get an Athlon up to 2.49GHz for comparison. Given the nearly clock-for-clock comparison with the Pentium 4 (actually, the Pentium 4 has a 48MHz advantage, but the comparison is close), I would have expected much higher scores from the Athlon. However, it seems that at higher clock speeds, the Pentium 4 is able to keep up much better with the Athlon than it has in the past.

At 1024x768 video resolution and with no anti-aliasing in Quake III, the Pentium 4 is actually still able to outperform the Athlon by about 4%. When you turn on anti-aliasing and anisotropic filtering, the test seems to be entirely video card limited, as the difference between CPUs is essentially nil.

In 3DMark 2002, the Athlon takes a 3.5% lead in the low resolution, but once again, when you turn on the high end features, the difference becomes negligible.

Jedi Knight is a little different. The Athlon maintains a 7-8% lead no matter if the high end features are turned on or not, nor whether the resolution is turned up or down. Further, in Comanche 4, the Athlon has a 10% advantage with features off, but only 3% with features on. In the final benchmark, Code Creatures, the Athlon has a negligible advantage, unless both features and resolution are turned to high settings.

Surprisingly enough, with the exception of the last, the Athlon couldn't boast more than a 10% advantage in any one benchmark, and for the most part, the advantage was much less. This is in sharp contrast to earlier benchmarks where an equal clocked comparison between the two architectures would show much of a distinction. But in this case, benefits of <10% are barely recognizable to the end user, showing the Athlon architecture to be much less compelling than it once was. Rather, it seems that a faster front side bus, more cache, and higher frequencies allow the Pentium 4 to finally hold its own against an equally clocked Athlon. At the end of the day, therefore, I would expect the Pentium 4 architecture to provide much higher frequencies overall, thus offering higher performance overall.

IMO, this speaks volumes for AMD's need to embellish their architecture, since the Pentium 4 seems to be scaling so much better in performance. Extra cache and a faster front side bus would surely help their K7 line, since the CPU seems to be much more bandwidth starved than before. Further, the lower latency integrated memory controller in Hammer should give a benefit as well. How much benefit, though, is still up to debate. With the Athlon architecture scaling so much less than expected, it's hard to say what future improvements may alleviate. Just speculation, though.

wbmw
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext