5.0 Introduction to testing sessions

The Radeon X800 Pro performances was compared with the NVIDIA Geforce 6800 GT and the ATI Radeon 9800 XT ones, both of them with a 256Mb memory and according to the reference design. These two cards seem to be the best to be compared with the Radeon X800 Pro: the first one seems to be an interesting alternative to the Radeon X800, the second one is a very good card that was the top among Radeon based cards and from which the X800 is the natural evolution.

The cards was compared in videogaming, the most interesting working field for this type of devices. For this reason in our measuring sessions we chose software able to analyse and bring out the performance response with the hardware features mostly used by games (synthetic benchmarks). There are, as well, games you can run as benchmark, using internal commands or third part utilities: we used some of them too in our testing.

Usually an output result strongly depends on many different conditions, and often a measurer can't control them unless making a statistical analysis of the results. For this reason we arranged to stabilize the motherboard power supply and to make our measuring in a uniform temperature environment in order to don't change the components' working point faking the results. We ran every test ten times, using the same graphic controller, then we calculated the arithmetic mean on the partial results. The mean is considered as the final result. We will show it without indicating the relative or absolute uncertainty.

5.1 Results understanding

We considered two parameters in our testing: the frames' number mean in a unit of time (FPS) and the total score given by the synthetic benchmarks. The FPS is very significant because is a game quality index as it is strictly connected with the games' running speed.

Using the card driver control panel we did the tests with two standard configurations, based on the testing session:

• Default layout with V-Sync off;

• Default layout with antialiasing filtering on (4X) and anisotropic filtering on (8X).

The first layout, together with a high graphic quality game setup, allows to best perform the hardware performances in terms of "smoothness" without damaging the visual quality; the second one, instead, gives a significant index for estimating the games' running capacity, using setups oriented to a high graphic output.

Hardware outline of the measuring system

• Motherboard Abit NF7 (NVIDIA nForce 2 Ultra);

• 1Gb Ram DDR400;

• ATI Radeon X800 PRO 256MB, 8x AGP, 472.5MHz/905MHz;

• NVIDIA GeForce 6800 GT 256MB, 8x AGP, 350MHz/1000MHz;

• ATI Radeon 9800 XT, 8x AGP, 412MHz/730MHz;

• Enermax EG465P Power Supply (35A/3.3V, 35A/5V, 33A/12V).

Software outline of the measuring system

• Microsoft Windows XP Pro with SP2;

• Microsoft DirectX 9.0c;

• ATI Catalyst 5.7;

• NVIDIA ForceWare 77.77.

Benchmark tools

• 3DMark05;

• 3DMark03;

• 3DMark2001SE;

• Unreal Tournament 2004;

• Doom3;

• Far Cry;

• Half-Life 2 (Counter-Strike: Source Beta).

You can download the 3DMark series tools from our video devices testing area, while you can find the listed games on the official home pages or on many video gaming web resources.

In the next page we compare the technical specifications of these three boards.

Visual Inspection

1 2 3 4 5 6 7 8 9 10 11

Video Cards Comparison

[ Go Back ]