The Test

Below, you can see our test rig configuration.

 Performance Test Configuration
Processor(s): AMD Athlon 64 3800+ (130nm, 2.4GHz, 512KB L2 Cache)
RAM: 2 x 512MB Mushkin PC-3200 CL2 (400MHz)
Motherboard(s): MSI K8T Neo2 (Socket 939)
Memory Timings: Default
Hard Drives: Seagate 7200.7 120GB SATA
Video Cards: GeForce 6800 Ultra 256MB
GeForce 6800 128MB
GeForceFX 5950 Ultra 256MB
GeForceFX 5900 Ultra 128MB
GeForce 5900 128MB
GeForceFX 5700 Ultra
GeForceFX 5600XT
Radeon X800 Pro 256MB
Radeon 9800XT 256MB
Radeon 9700 Pro 128MB
Radeon 9600XT 128MB
Operating System(s): SuSE 9.1 Professional
kernel 2.6.8-14-default
Driver: (ATI) SuSE 9.1 Supplement fglrx 3.12.0
NVIDIA 1.0-6111

You may have noticed that we are running an extremely new version of the Linux kernel, and very new ATI and NVIDIA drivers as well. For all intents and purposes, we are running a completely default SuSE 9.1 Professional install with the SuSE 9.2-RC3 kernel and brand new drivers. This was not an easy accomplishment, but was unfortunately the only manner in which we could install a platform compatible with both ATI and NVIDIA video cards on the Socket 939 architecture.

Our testing procedure is very simple. We take our various video cards and run respective time demos while using our AnandTech FrameGetter tool. We rely on in-game benchmarks for some of our tests as well - since FG will not run on Wine games. We post the average frames per second scores calculated by the utility. Remember, FG calculates the frames per second every second, but it also tells us the time that our demo ran, and how many frames it took. This average is posted for most benchmarks.

However, when testing our games, we find that some interesting patterns sometimes occur. For these instances, we have specially crafted the FG program to record our timedemo by taking the frames per second every second and dumping this data into a text file. We explained this in our initial FG announcement. Some graphs, particularly Wolfenstein and Unreal Tournament, have particularly fascinating trends, which we explore more in the evaluation.

All of our benchmarks are run three times and the highest scores obtained are taken - and as a general trend, the highest score is usually the second or third pass at the timedemo. Why don't we take the median values and standard deviation? For one, IO bottlenecks tend to occur due to the hard drive and memory, even though they "theoretically" should behave the same every time that we run the program. Memory hogs like UT2004, which tend to also load a lot of data off the hard drive, are notorious for behaving strangely on the first few passes.

Since we had issues with the ATI driver running Anisotropic Filtering, we did not run any tests with AF on. However, many of our games have sets of benchmarks with 4X Anti Aliasing disabled and enabled. At the end of this analysis, we also have a small section showing some of the differences with the various AA and anisotropic filters enabled.

More Configuration Unreal Tournament 2004 32-bit


View All Comments

  • - Saturday, October 24, 2009 - link">
    (air jordan, air max, shox tn, rift, puma, dunk sb, adidas) nike jordan shoes 1-24 $32
    lv, coach, chane bag $35
    COOGI(jeans, tshirts, hoody, jacket) $30
    christian audigier(jeans, tshirts, hoody) $13
    edhardy(shoes, tshirts, jeans, caps, watche, handbag) $25
    Armani(jeans, tshirts,) $24
    AF(jeans, coat, hoody, sweater, tshirts)Abercrombie & Fitch $31">
  • TheWounded - Monday, November 01, 2004 - link

    Its a nice test but i would have loved to see how the XGI volari cards would have done.
    I'm interested if the volari's could be a good choice for linux gamers. But unfortunatly there are no linux benchmarks involving the volaris.
  • henca - Thursday, October 07, 2004 - link

    This was a very nice comparision of mid- and high-end cards. It would be interesting to also see a comparision with low-end cards like Matrox G550, Intel Extreme graphics and the Radeon 9200 family.

    The good news about these cards is that they are all supported by the opensource DRI drivers. An up-to-date Linux distribution should support them out of the box without having to download and install any binary drivers.
  • MNKyDeth - Tuesday, October 05, 2004 - link

    I am a Linux gamer only so a benchmark comparison like this is great. I really enjoyed reading it. But, imo, there was a lack of games included in the benchmark roundup. I would like to see Savage, NWN, and either quake3 or Heretic 2 shown aswell.

    I also do not like the showing of wineX (Cedega) benchmarks as it defeats the purpose the gaming on linux. The only way I could recomend anyone to use wineX (Cedega) is if they don't own a copy of windows. If you do own a copy of windows do not use wineX for pete's sake, just dual boot, it is the better emulator after all.
  • jerrysiebe - Tuesday, October 05, 2004 - link

    For anisotropic filtering, I did a strings search in libGL and came up with something.

    >strings /usr/lib/ | grep ANISO

    Setting that, I can see a visible difference and get a FPS hit, so I believe it works. On my GF4 4200, I can set __GL_LOG_MAX_ANISO to 1, 2, and 4 and see the difference. Set to anything else I get no anisotropic filtering.
  • Thetargos - Monday, October 04, 2004 - link

    Excellent article, just a comment on the NVIDIA uninstaller... it plainly doesn't work as it should. The prlblem is that it substitutes (like the ATi driver) some libraries in the system, but unlike ATi's driver, NVIDIA's driver also makes a change in one library used for the Direct Redering Infrastructure, libdri.a specifically. So uninstalling the drivers with NVIDIA's uninstaller this won't be reverted (re-install of the XFree86 package or Xorg package is required, note only the core package is need).
    In favor of ATi's driver, the uninstallation is much easier and the system is restored to its previous stage, restoring the backup copy of that is the only system library it overwrites.
  • plamalice - Monday, October 04, 2004 - link

    The Nvidia AGPgart driver is causing problems with ATI cards (perhaps other non-nVidia card as well) on both Win and Linux when used on an nForce based mobo (of course). Nforce3 (150, pro150) have both caused me problems when using an ATI card until the gart driver was uninstalled.

    A poor attempt by nVidia to make ATI card appear unstable ? :P

    Anyways, if you have an nForce-based motherboard and an ATI gfx card, do not use nvidia's gart driver.
  • KristopherKubicki - Monday, October 04, 2004 - link

    directedition: i just symlink /mnt/cdrom to /media/dvdrecorder

  • mczak - Monday, October 04, 2004 - link

    "Keep in mind that we even run SuSE, a RPM derivative - not too different from Red Hat."
    That really doesn't make sense. RPM is just the package manager! If a dos version which uses rpm would exist, would you say that it is "not too different" too?

    "Below, you can see a screen grab from our ATI frame buffer playing Unreal Tournament at 800x600. The image should not be surrounded by a black border, but rather, stretched to the limits of the screen."
    This looks to me like you did not have configured 800x600 resolution in the Xfree config file (Sax2 will happily do that) - you cannot switch to fullscreen resolutions not configured usually with XFree/Xorg (though maybe the nvidia driver doesn't care).

    btw about aniso not working: I guess you could do that quite easily with framegetter? Just intercept the filter setting calls and replace them?
  • mczak - Monday, October 04, 2004 - link

    "On our MSI nForce3 board, this should have been the nvidia_agp module. However, try as we could, we could not get nvidia_agp and fglrx to play well with each other."
    This is a mistake, you do not need (and it will not work) the nvidia-agp module. For all A64 based boards, no matter if the chipset is from sis, via, nvidia or someone else, you need the amd64-agp module instead. It might have just worked with that - suse 9.1 loads it automatically for K8T800 chipset, but I think for some reason it doesn't get automatically loaded for nforce3 chipsets. It might have just worked loading it manually, saving you some time :-).

    "We are not entirely sure why, but even after completely removing the NVIDIA kernel module, we still had persistent errors installing the ATI drivers correctly."
    Removing the kernel module will do nothing. Nvidia drivers replace some of XFree/Xorg libraries, which are incompatible (I think libglx.a is affected by that, but there might be more), and ATI does not have its own version of these files. Uninstalling the nvidia driver with its own installer (which has an uninstall option) should get the original version back in place afaik.

Log in

Don't have an account? Sign up now