|
Test Setup
All tests were run on a clean installation of Windows ME with all drivers installed. No other programs were left running in the background while tests were running, no screen savers or power saving features were configured and the mouse was left still for the entire duration of the benchmark.
Since the processor that I had in this test system is a low-mid-level processor, the AMD Duron 650MHz, I felt that using the PowerColor PowerGene GeForce2 MX was sufficient enough for these tests. Regardless of which, because all tests were run at 640x480, using the MX's big brother, the GTS, would not have affected the scores much, if not, at all.
Benchmark Results
I would like to mention, once again, that I ran all benchmarks at a low resolution of 640x480 to avoid running into fill-rate bottlenecks on the GeForce2 MX card that was running the graphics subsystem during our tests.
|
SY-K7VTA
Pre-production |
SY-K7VTA
Final production |
| 3DMark2000 (640x480x16) - 3DMarks |
5624 |
5866 |
| SiSoft Sandra 2000 |
|
|
| CPU Dhrystone |
1950 |
2004 |
| FPU Whetstone |
885 |
905 |
| Integer MMX - it/sec |
2176 |
2206 |
| Floating-Point SSE - it/sec |
3003 |
3036 |
| CPU Memory Bandwidth MB/s |
361 |
404 |
| FPU Memory Bandwidth MB/s |
398 |
475 |
| Unreal Tournament (wicked400) - FPS |
26.04 |
26.27 |
| Quake III Arena (demo001) - FPS |
101.7 |
105.0 |
Well, I expect that most of you would have wanted a colorfull graph with all the eyecandy, but with only 2 products to make comparison, a graph would have confused most instead of making things easier.
A quick glance through the chart shows the entire story; SOYO has tweaked a whole lot out of their SY-K7VTA since the last time we had a chance to take a look a look at it. The production version managed 200 more 3DMarks, almost 4 more FPS and, what was the most improved, managed 77 more MB/s in SiSoft Sandra's FPU memory bandwidth test than the pre-production version we reviewed earlier. A look at those improvements clearly states how much more juice out of their design.
 On to: Conclusion
|
