Dual Benchmarks In Windows 2000
| OpenGL Performance | Quake 3 Arena "Demo 1" |
| 3D-Rendering | Cinema 4D XL R7 |
| 3D-Rendering | 3D Studio Max 4.2 |
| 3D-Rendering | Lightwave 7 |
| Audio-Encoding MP3 | Lame MP3 Encoder |
| Video-Encoding MPEG-4 | XMpeg 4.2a und Divx 4.2 |
| Office Performance | Sysmark 2001 |
| Linux Kernel Compiling | Suse Linux 7.3 (Kernel 2.4.13) |
| SiSoft Sandra 2001 | CPU and Multimedia Bench |
Because of the automatic resource management integrated into Windows XP, we have used Windows 2000 to test the dual boards in question - the only reason being that Windows XP optimizes the running of applications in the background. That means that we cannot exactly calculate benchmarks with these nuances.
All told, we ran 13 different benchmarking tests to present a multi-faceted and balanced picture of the performance of the two dual systems. In addition, this shows the suitability of each piece of software for multiprocessor use.
We determined OpenGL performance with the Quake III Arena benchmark. The various MPEG encoding benchmarks make for a comprehensive testing scenario: using Lame MP3 encoder, a 178 MB WAV file is coded in "MPEG-1 Layer 3 Format." The settings for this are as follows: variable bit rate from 0 to 256 kBit and optimum quality. We also examined the encoding of an MPEG-4 file from a commercial DVD. For some time now, we have also been using Newtek's Lightwave professional package, version 7b, to calculate rendering performance. The compilation of the latest Linux kernel has long been part of the standard repertoire. Sysmark-Benchmark 2001 is used to determine the office performance in dual operation. The main feature of this dual test is rendering performance. We used 3D Studio Max 4.2 and Cinema 4D XL R7 to test it.
