Benchmarks I
Benchmarking of the various combinations begins with Pifast. It simply calculates the constant Pi to the desired number of places. 10 million is chosen for this test. If you want to run it for yourself, click here for the benchmark standings and download link. Just unzip and click on the .bat file.
As both motherboards are running the XP2700 at 2171.5MHz, and both use the same timings, the results with a discrete Radeon 9800 are nigh-on identical. Using the on-board graphics exacts a very small toll if we're running in the abundant bandwidth dual channel mode. Single channel mode, however, has to juggle both the on-board graphics and CPU's memory requirements, thereby delivering a slower time.
Next we'll turn our attention to MP3 encoding. We're benchmarking by encoding a 610MB custom WAV file (U2's Pop album, incidentally) into 192kb/s MP3 using the LAME 3.91 encoder and Razor-Lame 1.15 front-end.
Only the single channel, on-board graphics combination lags behind to any significant degree.
DivX encoding, like Pifast, is a lover of bandwidth. The first VOB of American History X on a single-pass quality basis using DivX5.03.
It used to be the case that on-board graphics exacted a heavy toll on performance in 2D applications; sharing bandwidth has that effect. Dual channel DDR seems to be the perfect answer for those looking to run integrated graphics with a minimal performance decreases in non-3D applications.
Let's see how they fare in our SETI benchmark. A 0.417WU makes it an average unit. Time is in hours, minutes and seconds.
AMD-based machines have always been strong at mathematical tasks. The nForce2 and XP2700 combination shows its strength here.
