Review: ATI Radeon X1950 XTX

Back to Page 1

Test System Setup

The graphics boards tested for this article were installed in an Antec Sonata midtower case and the testbed built using the following components:

• Intel Core 2 Extreme X6800 (2.93 GHz) processor
• Intel 975XBX motherboard
• 2 GB (2 x 1 GB) of Corsair DDR2 800 memory (4-4-4-12)
• WD Raptor 150GB hard drive
• Plextor PX-712S DVD drive
• Creative SoundBlaster X-Fi sound card
• Enermax 550 Watt ATX PSU
• Dell 2405FPW 24” widescreen LCD

We decided to pit the Radeon X1950 XT against the X1900 XTX to examine what performance improvements the refresh part and its rather significant memory bandwidth increase would display in our benchmark suite. The latest BIOS version and chipset drivers were installed on the test system, along with the Catalyst 6.9s. For all conducted tests, Catalyst A.I. was placed at Standard and the Mipmap detail option placed at high quality in the control panel. The licensed version of Fraps 2.7.4 was used to record benchmark scores for applications that lacked the option to record their own frame rates. Testing was conducted using the following software:

• 3DMark06 (v1.0.2)
• IL-2: Pacific Fighters (v4.04)
• MS Flight Simulator 2004 (v9.1)
• A Company of Heroes (v1.2)
• Ghost Recon Advanced Warfighter (v1.21)
• Call of Duty 2 (v1.3)
• rFactor (v1150)

When possible, application anti-aliasing was used rather than forced via the driver control panel; also anisotropic filtering was not tested with its High Quality (angle independent) mode enabled. Each game title was tested at 1024x768, 1280x1024, 1600x1200, and 1920x1200 with 4x anti-aliasing and 8x anisotropic filtering. The resolution of 1920x1200 was used since it is the highest display option available with the test rig’s 2405FPW display.

Benchmark Test Scores

We included Futuremark’s 3DMark06 see what performance delta would exist between the aggregate score of each shader model’s set of included tests at the highest available resolution (1920x1200) with 4x anti-aliasing and 8x anisotropic filtering both enabled in the application’s options menu.

The X1950 outperforms the older board in both shader model test sets by roughly 100 points. This difference is more significant in the SM3.0 set, but neither delta represents even a 7% increase. Futuremark has coded 3DMark06 to be more shader-intensive in its game tests, and the increased bandwidth of the 1950 clearly isn’t of much assistance in this particular synthetic set of tests.

Go To Page 3

Click here to go to top of this page.