Gaming Tests Methodology
These games and applications were used as benchmarks:
First-Person 3D Shooters
- Battlefield 2
- The Chronicles of Riddick
- Call of Duty 2
- Doom 3
- Far Cry
- F.E.A.R.
- Half-Life 2
- Project Snowblind
- Quake 4
- Serious Sam 2
- Unreal Tournament 2004
Third-Person 3D Shooters
- Prince of Persia: Warrior Within
- Splinter Cell: Chaos Theory
Simulators
- Colin McRae Rally 2005
- Pacific Fighters
Strategies
- Age of Empires 3
- Warhammer 40000: Dawn of War
Semi-synthetic benchmarks
- Aquamark 3
- Final Fantasy XI Official Benchmark 3
Synthetic benchmarks
- Futuremark 3DMark03 build 360
- Futuremark 3DMark05 build 120
S3, ATI and NVIDIA drivers were set as follows:
S3 Chrome 6.14.10.2133-15.17.14h:
- Wait for monitor sync: Off
- Fast trilinear filtering: Auto (only for Direct3D)
- Fast destination clear: Auto (only for Direct3D)
- Fast Z clear: Auto (only for Direct3D)
- Detail level adjust: 0 (Quality)
- Max depth buffer size: By app (only for Direct3D)
ATI CATALYST 5.12:
- CATALYST A.I.: Standard
- Mipmap Detail Level: Quality
- Wait for vertical refresh: Always off
- Adaptive antialiasing: Off
- Temporal antialiasing: Off
- Quality AF: Off
- Other settings: default
NVIDIA ForceWare 81.95:
- Image Settings: Quality
- Vertical sync: Off
- Trilinear optimization: On
- Anisotropic mip filter optimization: Off
- Anisotropic sample optimization: On
- Gamma correct antialiasing: On
- Transparency antialiasing: Off
- Other settings: default
We chose the maximum graphics quality settings in games for all the graphics cards and the maximum possible rendering mode, too: Shader Model 3.0 for GeForce 6 and RADEON X1000 architectures and Shader Model 2.0 for S3 Chrome and RADEON X700. We used the in-game benchmarking tools to record and reproduce a demo and measure the reproduction speed in frames per second, if possible. If not, we used the FRAPS utility. We also measured minimal as well as average fps rates whenever possible.
We turned on 4x FSAA and 16x anisotropic filtering for the “eye candy?test mode using the game’s menu or from the graphics card driver. If the game engine didn’t support FSAA, we didn’t test the cards in the “eye candy?mode.
The following graphics cards were selected as opponents to the S3 Chrome S27 in gaming tests:
- RADEON X1300 PRO (RV515, 600/800MHz, 4pp, 2vp, 128-bit, 256MB)
for details see our article called ATI RADEON X1300 PRO: Entry-Level Gets a Decent Performer? - RADEON X700 (RV410, 400/700MHz, 8pp, 6vp, 128-bit, 256MB)
- RADEON X300 (RV380, 325/400MHz, 4pp, 2vp, 128-bit, 128MB)
- GeForce 6600 (NV43, 300/500MHz, 8pp, 3vp, 128-bit, 128MB)
- GeForce 6600 GDDR2 (NV43, 350/800MHz, 8pp, 3vp, 128-bit, 128MB)
for details see our article called XFX GeForce 6600 GDDR2: GeForce 6600 Rejuvenated - GeForce 6200 (NV43, 300/500MHz, 4pp, 3vp, 128-bit, 128MB)
The RADEON X700, RADEON X300 and GeForce 6600 graphics cards were emulated by down-clocking the RADEON X700 PRO, RADEON X600 XT and GeForce 6600 GT, respectively.
