Since the application based scores above do seem to be directly linked to CPU speed, lets look at the actual CPU scores with Winbench 99 CPUmark99 and 3D Winbench 2000 CPU scores.
Both CPU tests do support the application based benchmarks results that CPU performance is not being limited by other sub-systems. It is interesting to note that the Duron scores look to be a bit ‘flat’ compared to the Athlon. There is a larger step from the Duron 950 to the Athlon 1.0 than between the individual results for CPU’s of the same type. Again, due to the smaller L2 cache or lower FSB of the Duron. Also there is only a 150MHz spread between the two Duron’s and with the Athlon’s it was 200MHz.
Will graphics show the same trend? Lets look at the Winbench graphics tests that are application based, as well as two game based:
Nice progressive graph, not limited in application based graphics.
Here we again see a larger step between the Duron and Athlon. Also the difference between the Athlon 1.4 and 1.2 is 7 points but between the Athlon 1.2 and 1.0 is 14+ points. Looks like we are becoming limited by something other than the CPU.
And finally Video 2000 shows a progressive scale, but does taper off as the speed increases. The Athlon again seems to have an advantage over the Duron, but is it due to the Athlon’s larger L2 cache or higher FSB?
Be the first to discuss this article!