The choice of competeing cards for this review was fairly straight forward. NVIDIA's 7900 GS which comes in between $199-$249 is ATI's main target competition for the X1950 Pro, and with a new MSRP of $195 ATI's own X1900 GT also makes a fair comparison point. We're using the "old" X1900 GT however so keep that in mind. The "new" X1900 GT with the different core and memory clock speeds, the one you will probably find in retail now, performs slightly different.

We've also throw in an ATI X1950 XTX and BFG 7900 GTX OC for comparison points against X1950 Pro Crossfire. Two X1950 Pro cards will set you back about the same as one X1950 XTX or one 7900 GTX so we feel that comparison is legitimate.

Here are more details about our test system and methods:

Games Benchmarks (click for settings)

• Prey, v1.2 (OpenGL)
  • Video
  • Texture Quality: Highest Quality
  • Advanced 1
  • Use Shader Programs: Yes
  • Shader detail: Highest
  • Enable Shadows: Yes
  • Enable Specular: Yes
  • Enable Bump Maps: Yes
  • Vertical Sync: No
  • Anti Aliasing: Various
  • High Quality Specular: Yes
  • Sharpen Bumpmaps: Yes
  • Enable Bump Maps: Yes
  • Advanced 2
  • Use Glow: Yes
  • High Particle Detail: Yes
  • High Quality Skinning: Yes
  • Image Anisotropy: Various
• Half-Life 2, Source Engine 7 build 2953 (DirectX)
  • Model Detail: High
  • Texture Detail: High
  • Shader Detail: High
  • Water Detail: Reflect All
  • Shadow Detail: High
  • Antialiasing mode: Various
  • Filtering mode: Various
  • Wait for vertical sync: Disabled
  • High Dynamic Range: None
• F.E.A.R., v1.07 (DirectX)
  • Computer: Maximum
  • Graphics card: Maximum
    • FSAA: Various
    • Light Detail: Maximum
    • Enable Shadows: On
    • Shadow Detail: Maximum
    • Soft Shadows: Off
    • Texture Filtering: Various
    • Texture Resolution: Maximum
    • Videos: Maximum
    • Pixel Doubling: Off
    • DX8 Shaders: Off
    • Screen Resolution: Various
    • Shaders: Maximum
• Battlefield 2, v1.4 (DirectX)
  • Terrain: High
  • Effects: High
  • Geometry: High
  • Texture: High
  • Lighting: High
  • Dynamic Shadows: High
  • Dynamic Light: High
  • Anti-Aliasing: Various
  • Texture Filtering: High
  • View Distance Scale: Max
• Quake 4, v1.3 (OpenGL)
  • Video Quality: Ultra Quality
  • High Quality Special Effects: Yes
  • Enable shadows: Yes
  • Enable specular: Yes
  • Enable bump maps: Yes
  • Vertical sync: No
  • Antialiasing: Various
  • Multiple CPU/Core: Yes
• Half-Life 2: Lost Coast, Source Engine 7 build 2953 (DirectX)
  • Model Detail: High
  • Texture Detail: High
  • Shader Detail: High
  • Water Detail: Reflect All
  • Shadow Detail: High
  • Antialiasing mode: Various
  • Filtering mode: Various
  • Wait for vertical sync: Disabled
  • High Dynamic Range: Full
• Splinter Cell: Chaos Theory, v1.05 (DirectX)
  • Shader Model Type: Various
  • Shader Options (SM3.0 only)
    • HDR Rendering: On
      • Tone Mapping: On
    • Parallax Mapping: On
    • High Quality Soft Shadows: On
  • Shadow Resolution: High
  • Trilinear Filtering: On
  • Specular Lighting: On
  • High Quality Textures: On
  • Vertical Sync: Off
  • Hardware Shadow Mapping: On
    
• Far Cry, v1.4 (DirectX)
  • Texture Quality: Very high
  • Texture Filter Quality: Trilinear
  • Anisotropic filtering level: Various
  • Particle Count: Very High
  • Special Effects Quality: Very High
  • Environment Quality: Very High
  • Shadow Quality: Very High
  • Water Quality: Ultra High
  • Lighting Quality: Very High

Resolutions

• 1280x1024 / 1280x960
• 1600x1200
• 1920x1200 / 1920x1440

Image Quality Settings

• No AA & No AF
• 4x AA & 16x AF

Test Systems	NVIDIA	ATI
Motherboard ( chipset )	Foxconn C51XEM2AA (NForce 590 SLI)	MSI K9A Platinum (ATI RD580/SB600)
CPU	AMD Athlon FX-62 @ 2.8GHz
Graphics Card	NVIDIA 7900 GS PNY Verto 7900 GS for SLI BFG 7900 GTX OC	ATI Radeon X1950 Pro x2 HIS Radeon X1900 GT x2* ATI Radeon X1950 XTX*
Driver Version	Forceware 91.47 WHQL	ATI 8.291 Beta Catalyst 6.10 Beta*
Memory ( Timings )	2GB (2x1024MB) SuperTalent PC2-6400 DDR2 @ 800MHZ ( 4-4-3-8 2T)
Hard Disk	Western Digital Caviar WD2500KS SE16 250GB
Sound	Onboard
Network	Onboard
PSU	PC Power & Cooling 1KW Turbo-Cool Quad-SLI
OS	Windows XP Pro SP2, DX9c (August 2006)

Card Specifications	ATI X1950 Pro	HIS X1900 GT	NVIDIA 7900 GS	ATI X1950 XTX	BFG 7900 GTX OC
Core	RV570	R580	G71	R580+	G71
Silicon Process	80nm	90nm	90nm	90nm	90nm
Transistor Count (millions)	330	384	278	384	278
Core Speed MHz	575	575	450	650	670
Memory Speed MHz (Effective)	690 (1.38GHz)	600 (1.2GHz)	660 (1.32GHz)	1,000 (2.0GHz)	820 (1.64GHz)
Memory Size	256 MB	256 MB	256 MB	512 MB	512 MB
Bus Standard	PEG x16	PEG x16	PEG 16x	PEG x16	PEG 16x
Bus Width	256bit	256bit	256bit	256bit	256bit
ROPs	12	12	16	16	16
Pixel Shaders	36	36	20	48	24
Vertex Shaders	8	8	7	8	8
Peak Memory Bandwidth (GB/s)	44.2	38.4	42.2	64.0	51.2
Pixel Fillrate (million pixels/sec)	6,900	6,900	7,200	10,400	10,720
Texel Fillrate (million texels/sec)	6,900	6,900	9,000	10,400	16,080
API Compliancy	DX 9.0c	DX 9.0c	DX 9.0c	DX 9.0c	DX 9.0c

The Windows XP desktop was set to 1280x960 with a 32bit color depth and 85Hz refresh rate for all tests. Refresh rate locks for 3D graphics modes, as supported by both NVIDIA and ATI graphics control panels, was not enabled. V-Sync was forced off via the graphics card control panel as well. For the NVIDIA cards image quality was set to "High Quality" in order to reduce texture shimmer. We feel enabling this setting makes for a more legitimate comparison when comparing NVIDIA cards to ATI cards.

All graphics card control panel settings for the ATI cards were left to their default settings.

Anti-Aliasing and Anisotropy were applied in the game engine where the options existed. For games that did not support those options natively, the graphics card control panel was used.

Custom batch files were used when possible for automated benchmarking, which are available upon request. When manual benchmarking was necessary Fraps was used.

Benchmarking was done with Windows set to the "Adjust for best performance" profile and all unnecessary Windows services and hardware devices were disabled. The latest drivers for each necessary hardware component were installed prior to testing and kept consistent throughout.

Sound and networking interfaces were enabled for all tests.

Windows XP was installed fresh on two seperate but identical hard-drives prior to testing (one for the ATI system, one for the NVIDIA system).