Navigate: Advice Beginners BIOS Guide CPUs Links Mainboards Memory Network Storage Video/Sound Cards Contact Forum SiteMap Sponsors WebNews Home	.	.
Prices: Mainboards ABIT ASUS Chaintech Shuttle Soyo Tyan CPU Intel P4 2.4C-800 P4 2.6C-800 P4 2.8C-800 P4 3.0-800 P4 3.2-800 AMD AthlonXP XP 1700+ XP 2000+ XP 2400+ XP 2500+ XP 2700+ XP 3000+ XP 3200+ Athlon64 Athlon64 3200+ Athlon64 FX-51 Opteron Opteron 240 Opteron 242 Opteron 244 Opteron 246 Memory Corsair Crucial Kingston Mushkin OCZ Search Prices:

LOSTCIRCUITS	SHORTCUTS: New Performance Dimensions Test Configuration Cachemem Sandra Everest Analysis-1 FarCry The Poodle's Core More Benchmarks Conclusion Give Us Some Feedback to Help Us Improve our Reviews
Intel Pentium4 Prescott / 925XE chipset New Dimensions in Performance
(Review by MS November 21, 2004)

Data Analysis

To get to the bottom of what we saw, we took some dozen data points and plotted the numbers against each other, that is, the frequency shown against the returned benchmark result:

Measured bandwidth (blue) vs memory frequency (red). For better visibility, the freqency values were multiplied with a factor of 10, e.g. 1670 represents 167 MHz memory data rate as shown by Everest.

The same trend is seen for write transactions.

Measured bandwidth (blue) vs memory frequency (red). For details see above.

For further analysis, we plotted a calibrated two lines-plot

In order to get a better visual fit of the data, we plotted the different READ points with the highest and lowest numbers set a maximum and minimum of the two scales, respectively. The two graphs are almost mirror-symmetric to each other, except that the "167 / 12008" point is somewhat off, which skews the plot.

To check whether the data fit into the same group we normalized them to a 533 MHz reference value. The result is the same as in the plots above, all data points fall into the same group with the exception of the last point at which a non-linearity is introduced that we will explain later.

next page: => What about Gaming =>