Gaming and 3D
For 3D performance, let’s first look at a synthetic benchmark that gives both an overall and CPU score: 3D Mark 2000.
Notice that the CPU score difference is about 9%, but the actual test score is only about 3% – Why? Pretty simple (I believe), the video card used is pretty much maxed out because FPS scores (not shown here) don’t really change at all with the different settings.
Quake III is always a good test to see differences in memory speed. I’ve never figured out if its actually a good benchmark of gaming performance, but at least you can use it to ‘see’ changes in memory settings and types.
A difference of almost 20 FPS (remember, the highest score is using an overclocking speed) from the highest to the lowest is pretty significant. The question really is if it is a ‘true’ test of system performance that will be reflected in ‘Real World’ use, or is it just a quirk of the benchmark? Since we don’t see any other benchmark show a gain of almost 20% it is suspect, but then again it could be an indication of what kind of system usage would benefit most from using the most aggressive settings, and what usage will benefit the most from an overclocking speed.
Final reality is an older benchmark, not bad in itself, since if you use an older program it may profile your usage. Also, like Quake, it is sensitive to different memory speeds and settings.
No real gain here when using the 142MHz overclocking speed, but there is again a significant (the most of any test run) difference between the lowest and highest score. Notice how much of a difference there is between the individual test scores.
Be the first to discuss this article!