Navigate: Advice Beginners BIOS Guide CPUs Links Mainboards Memory Network Storage Video/Sound Cards Contact Forum SiteMap Sponsors WebNews Home	.	.
Prices: 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:

Aside from the physical resolution and color depth, latencies as we hinted earlier play a dominant role in the overall performance of UMA systems. In particular, the address and command decode latency settings can convert a race horse into a slug and back as we showed in this article. Again, the combination of latencies and resolution is the magic ingredient. In this article, we will only show the effect of changing the command rate from 2 to 1 cycle on the outcome of 3DMArk2001SE using the integrated graphics.

3DMark2001SE - ProSavage8

A 13% performance difference from changing a single latency setting. Keep in mind that this is not a difference in just frame rate but of the composite score of four different benchmarks.

For the record, all other benchmarks were run at a 1T CMD Rate. For gaming benchmarks we only show 3DMark2001SE using the integrated graphics. The reason is simply that the ProSavage8 is simply not made for this kind of task and quite a few benchmarks will simply not run on it. Comanche4 requires hardware T&L, CodeCreatures ..... we won't even go there. Suffice it to say that the 2D quality is not the greatest but it is not as bad as we have sometimes heard. Even at 1600 x 1200 resolution on a good monitor (Acer P911) there were no fuzzyness or shadows but don't expect Matrox or ATi quality either.

3DMark2001SE - Radeon 9700 Pro

The SK41 is slower than the A7V8X but the difference is not earthshaking.

Unreal Tournament (demo) Antalus Flyby 1280x1024x32

Something else is limiting here, either the CPU performance or else the graphics card. The lower memory speed does not appear to make any difference in this benchmark

Comanche4 800x600 x32

A minor difference between the two systems is all.

next page:    => Overclocking, Conclusion =>