I hope you don’t find these graphs confusing, I’m trying to prove a point by comparing the P4 2.0 using the i845 and i850 chipsets, and the Athlon MP 1.2GHz using the KT266 to both of those. Each set of bars shows the percent advantage (or disadvantage) in performance of the first named platform vs. the second named platform in each benchmark.
OK – So what are we looking at here? The bottom two bars show the P4/i845 performance compared to the Athlon/KT266 platform. Not too good for the P4, because even at 2.0Ghz it is about 5% slower than the AMD at 1.2GHz.
So what about if we compare the i845 to the i850 (middle two bars)? There is a big difference, as the i850 platform averages about 7% better performance than the i845 using the same CPU.
Since the Athlon/KT266 was faster than the P4/i845 platform, how does it fare when up against the P4 using the i850 chipset? Well the P4/i850 does beat Athlon (top two bars), but not by much, and that difference could easily be wiped out by either the new KT266A chipset or by using a higher speed (standard production part that will be available soon) Palomino core Athlon.
It looks to me that Intel may have something to worry about in the performance department, but let’s look at a few more test results before making a judgment.
A little bit different story here, as the P4 with either the i845 or i850 chipsets hold their own or even outperform the Athlon, but I wonder what those results will show when using a Palomino Athlon 1.5GHz and VIA KT266A chipset sometime next month? The i850 again shows an advantage over the i845 here.
It’s interesting to see how different the results are between Winstone and SYSmark, and how the different mix of programs used in the suites along with the different testing methods give such different results.
Be the first to discuss this article!