ATI Radeon X800 Pro 256 MB

Reviews: Hardware

Motherboards | Power Supplies | Memory | Systems | Video Cards

Author:
Editor:
Sponsor:
Published:

Brian Kristensen
Kurtis
ATI
Sep. 8, 2004

<< Previous

Next >>

Test Setup

Due to the release of Doom 3 and the Half-Life 2 Source Engine benchmarking tool, we have revised the list of games that we benchmark with. To test the performance of the different cards, we will be using six popular games: Far Cry, Need for Speed Underground, Splinter Cell, Call of Duty, and as I previously mentioned, Doom 3 and the Half-Life 2 benchmark. These games vary greatly in their use of shader effects and range from almost pure geometry (COD), to games where every surface has some sort of advanced shader effect (Doom 3), and everything in between.

The system we are using for benchmarking consists of an ABIT IC7-MAX3 motherboard which utilizes the Intel 875P chipset. An Intel Pentium 4 2.4c with 800 MHz FSB and Hyper-Threading technology is used along with 1 GB of Corsair PC3200 XL Pro memory (2-2-2-5 timings). Every component of the system is at stock or default speeds set by the manufacturer. While this definitely is not a top-notch system, it is sufficient for our purposes and if anything will result in benchmarking scores that the majority of our readers can relate to.

Windows XP was installed onto a freshly formatted Seagate Barracuda 160 GB SATA hard drive connected via on-chip SATA. All the latest Windows Updates were downloaded and installed except for Service Pack 2. The latest video drivers at the time of testing were installed - Catalyst 4.8 drivers for the ATI cards, Forceware 61.77 drivers for the NVIDIA cards.

In the video card control panel V-SYNC was disabled and the cards were set to "quality" over "performance." For the 9800 XT and X800 Pro that means the "Texture Preference" and "Mipmap Detail Level" were both set to "High Quality" while Anti Aliasing and Anisotropic Filtering were custom for each test. For NVIDIA cards, the "Image Settings" slider was set to "High Quality." For the 6800 GT, this automatically disables the Trilinear optimizations. However, the 5950 Ultra does not have the ability to disable the optimization. Everything else was left to the default settings.

Through our testing, we want the cards to display their best image quality as fast as they possibly can. This usually requires a few basic tweaks to the card's settings through the graphics control panel. However, we do not want to go too in depth and start tweaking variables that the average user wouldn't even touch. Our benchmarking results display how well each card performs when displaying the best image quality possible. Because some cards may support optimizations even at the highest quality settings, we compare the quality of the images each card produces on the "Image Quality" page.

To ensure the best possible results, we benchmark each game three times for each test and average the results. In most cases, we play a recorded time-demo which prints the average FPS (frames per second). However, some games differ in the way they record and play demos. Additionally, not every game supports demo or time-demo playback. More information and details regarding how each game benchmark is carried out can be found on the respective benchmarking pages.

Test Computer:

ABIT IC7-MAX3
Intel Pentium 4 2.4c
1 GB Corsair TwinX PC3200 XL Pro (Dual Channels; 400 MHz DDR @ 2-2-2-5)
Seagate Barracuda 160 GB SATA hard drive
Windows XP Professional SP1 w/ DirectX 9.0c and latest updates

When analyzing the performance graphs on the following pages, I will sometimes refer to settings that offer the best gaming experience. To me, a good gaming experience offers the right mixture of speed and image quality. Playing a game at 50 FPS with smooth edges and crisp textures would offer a better overall gaming experience than playing at 100 FPS with "jaggies' and noticeable texture filtering. Personally, I think 40-50 FPS allows for a smooth experience and my comments will stem from that belief. Your opinions may differ.

<< Previous

Page 4 of 13

Next >>

Page 1: Introduction
Page 2: The Card
Page 3: The Card (cont.) / The Bundle
Page 4: Test Setup
Page 5: Benchmarks: Doom 3
Page 6: Benchmarks: Half-Life 2
Page 7: Benchmarks: Far Cry
Page 8: Benchmarks: Need for Speed Underground
Page 9: Benchmarks: Splinter Cell
Page 10: Benchmarks: Call of Duty
Page 11: Image Quality
Page 12: Overclocking
Page 13: Conclusion

Subscribe to Video Cards [more info]

Add Comment

4 User Comments

1 - Posted by Rich on September 9, 2004 - 8:45 am

I am curious why you chose not to install SP2 for the testing. Any idea if there is a performance difference between SP1a and SP2 for XP?

2 - Posted by Brian on September 9, 2004 - 11:16 am

Testing of the cards began prior to the release of SP2. After the upcoming review of the XFX 6800 GT, the test system will have SP2 installed.

3 - Posted by Rich on September 10, 2004 - 8:42 am

Makes sense, I would be a bit interested to find out if there are any performance differences between SP1a and SP2.

4 - Posted by Brian on September 10, 2004 - 8:49 am

http://www.hardcoreware.net/reviews/review-237-1.h...

SP1 and SP2 seem to constantly switch places in different performance tests. Overall SP2 is slightly faster but definitely not noticeable unless you are comparing bench #s.

Add Comment

To add a comment without being a member, you may omit the password field, but you must enter your name (or nickname) along with your comment. * Denotes required fields.