Also, all benchmarks were performed with a single NX6800-TD128E as well as two of them in an SLI configuration.

You can use this percentage calculator to play around with the various differences between any combination of single, SLI, default, or overclocked speed settings.

Here is an example of the command line I used for the tests below:

doom3.exe +set logfile 1 +TimeDemoQuit viavga_d3_bench.demo

My autoexec.cfg looked like the example shown below and was of course changed per different resolution and AA settings.

I made use of the standard Doom 3 timedemo demo1. I also used the very stressful VIA Custom DOOM 3 Benchmarking Multiplayer Map timedemo.

Doom3 Benchmark Results - 0xAA  8XAF Athlon 64 3500+, MSI K8N Neo4/SLI, 1GB RAM Application Antialiasing and Anisotropic Filtering Results are in Averages

Doom3 Benchmark Results - 4xAA  8xAF Athlon 64 3500+, MSI K8N Neo4/SLI, 1GB RAM Application Antialiasing and Anisotropic Filtering Results are in Averages

My CPU bottleneck is very apparent at the lower resolutions at 0xAA. The overclocking benefits are really nice once you step up to 1600x1200 for a single card.

Here are some more pics that I know some of you will like to have a look at.

MSI NX6800-TD128E Thermal Tape Memory Pads

MSI NX6800-TD128E NV42 GPU

MSI NX6800-TD128E Hynix Memory Module

For those that would like more details on this particular part you can find out more details here.

Next Page: Half-Life 2 Benchmarks