Testing Methodology:

Motherboard:	MSI K8T-FIS2R Socket 754
CPU:	AMD Athlon 64 3000+
Memory:	Corsair PC3200 XL (2-2-2-6)
Power Supply:	OCZ PowerStream 600W
DVD-RW:	Plextor 712-A
Hard Drive:	Seagate Barracuda 7200.7 SATA
Contenders:	ATI X800 Pro (475/900) Chaintech AA6800 (358/770) Gigabyte 6800 (325/700) Albatron Trinity 6800 (325/700) Gainward 6800 (325/700) BFGTech 6800OC (350/700)
NVIDIA Drivers:	Forceware 66.93
ATI Drivers:	Cat. 4.12

We reinstalled Windows XP SP1 on our test bench before performing our performance tests. The hard drive was reformatted and windows updates were installed to keep our system up-to-date. We then upgraded our system from DirectX 9.0b to 9.0c version. Everything in the BIOS was set to default expect for memory timings and on-board LAN was disabled. We set the AGP Aperture size to the same amount as the video card’s physical memory. For Gigabyte’s 6800 and Chaintech’s AA6800, it was set to 128MB. And for ATI’s X800 Pro, it was set to 256MB. Please note the clock speeds of these cards again to clear any confusion you may have. Also note that Chaintech is retailing their AA6800 as a Turbo Edition model with a default 10 percent overclock, hence, the higher memory and GPU speeds. We decided to leave the clock speed of AA6800 in its retail condition for comparison.

AA and AF settings were changed globally through control panels and the image quality was set to "High" for all our tests. We decided to leave the games at their default settings for the most part; however, we did make some changes during certain points. Please refer to the list below for information on these changes:

Aquamark 3: We used standard test settings and a downloadable version for testing purposes. NVIDIA’s in-game optimizations were turned on only for Aquamark to make use of its 4xAF settings. For all other tests, in-games optimizations were turned off, but please understand that difference between in-game optimizations vs. "non-in-game" optimizations is far to none and it does not have much, if any, impact on the overall performance of the card.

Final Fantasy XI: We used Final Fantasy’s official benchmarking utility to perform our tests. Unfortunately, the utility only tests the card at either low or high quality settings with 1024x768 as the standard resolution. Surely enough we chose the high quality for our testing purposes. The score reported is not in frames per seconds, but it is an overall system score that will allow you to compare your system’s performance and make sure that it’s Final Fantasy XI ready, so to speak.

Need For Speed - Underground: We used a demo version of the game to make our testing consistent and controllable throughout all cards. Since NFSU only supports highest resolution of 1280x1024, we had no choice but to test the cards at this resolution with both AA and AF turned on and off. All settings were set to default expect the fog and blur features were turned off and the overall quality of the game was set to "High" specifications.

Far Cry: For this game, we used Ubisoft’s 1.3v patch for the "Training" demo. The image quality was set to "High" with no in-game AA settings turned on.

Halo: We used Halo’s built-in benchmarking demo for this performance analysis. We did not test the game with AA settings turned on because of the fact that Halo does not support AA functionality. This only left us with 8xAF settings, which we used at both 1280x1024 and 1600x1200 resolutions. Every other option in the game was set to default.

Unreal Tournament 2004: For this game, we used an open-source benchmarking utility, UMark, and BR-Colussus map for our testing. All game settings were set to default.

Doom III: We used Doom III’s Demo 1 for our testing at "High" settings. As you may or may not know, Doom III requires two runs of the demo to yield correct scores. The first time it cache the textures of the game while the second time it measures the performance more accurately with all the textures already pre-loaded. Obviously we used the pre-loaded, second run scores for our comparison purposes.

Half-Life 2: Our Half-Life 2 results are based on AnandTech’s Canal Demo. In-game Antialiasing was turned off, every setting was set to high, filtering mode was set to Trilinear and water detail was set to Reflect World. AA and AF settings were altered via drivers for global changes.

X2-The Threat: We utilized X2-The Threat’s benchmarking utility with the set resolutions. Automatic quality settings for the game were turned off and all quality settings were made globally from driver control panel (AA and AF settings).

Comanche 4: We decided to use this DirectX 8.1 compliant game for last generation game comparison. The game was tested at standard settings with no in-game AA enabled throughout any our tests. We used Comanche’s benchmarking utility for our regular comparisons.

Also, v-sync was turned off in all our tests and each test was run three times to confirm the numbers. The average of the three runs was taken for competitive analysis. We would like to state that we are waiting for Half-Life 2 to enter the market for us to use it in our standard video card test suite. As soon as the game enters retail channels, we will definitely include it in our future video card evaluations.