In keeping with our desire to refresh our GPU test suite periodically, we’re going to be redoing our GPU test suite to rotate in some more modern games, along with rotating in some DirectX11 games capable of taking advantage of this generation of GPU’s full capabilities. And while we already have a pretty solid idea of what we’re going to run, we wanted to throw out this question anyhow and see what responses we get.

What games would you like to see in our next GPU test suite, and why?

What we’d like to see is whether our choices line up with what our readers would like to see. We can’t promise that we’ll act on any specific responses, but we have our eyes and ears open to well-reasoned suggestions. So let us know what you think by commenting below.

Comments Locked

240 Comments

View All Comments

  • mindless1 - Thursday, March 18, 2010 - link

    I do not want to see benchmarks that only compare what each GPU has in FPS at the same game settings.

    NO, what is useful to the typical reader or potential buyer is, what, if any, quality settings are compromised to attain a playable framerate.

    If it just isn't possible to play at 20FPS (min) and at least 45FPS avg. I WANT to know that, but I don't care to know that a $200 card is faster than a $130 card at the same settings... I think we are past stating the obvious.

    So I want benchmarks that, if the above is too laborious, set an average framerate of 35FPS threshold, then note what needs disabled to attain that if it is possible at all.

    If it is not possible then the price of the card is pretty irrelevant isn't it? I mean, no matter how cheap it is, if it can't do the job at all there is no value in it for real world applications.
  • chadkellycolorado1 - Thursday, March 18, 2010 - link

    Don't leave out the growing croud of non-Microsoft gamers. Consider adding an OpenGL based game (open source with multi-platform support).
  • T2k - Thursday, March 18, 2010 - link

    It's a pretty well-written benchmark tool, with loads of option and customization options eg. switching between DX10/10.1/11 etc, capable to stress any high-end machine.
    It's using the latest X-Ray v1.6 engine of GSC Game World.
  • Browser - Wednesday, March 17, 2010 - link

    Engines not games.

    I think it would be a better idea to focus on engines rather than games, id Tech, CryEngine, Unreal, Gamebryo, Essence Engine, trying to get at least one of each game genre (RTS/FPS/etc) while covering the big boys.

    Also there are newer versions for some great engines due out in the near-ish future which you might want to consider waiting for or being ready to update to them, specifically the id Tech 5/6 and CryEngine 2/3

    Personally we don't need a gamut of resolutions tested but a good range focusing on the commonly used ones, even just 3 or 4 resolutions such as 1920x and below. Unrealistic or uncommon resolutions should be left to specific reviews, such as when comparing multi-monitor situations. The following might not be a bad idea, 1920x1080, 1680 x 1050, 1280x1024, as they are either common or high end.
  • Browser - Wednesday, March 17, 2010 - link

    (don't see edit)

    Someone mentioned OpenCL for it's physics which I have to say is a great idea, Anandtech could help publicising an open alternative to nvidia only closed PhysX. Since physics/OpenCL is both important to games and is affected by the graphics card I think it's a great idea.

    Something else I've wondered about, is a raw benchmark for the GPUs, showing FLOPS etc. It would give a nice idea of general power the GPUs have.

    These things could give you something over the other sites also benchmarking graphics cards.
  • tynopik - Tuesday, March 16, 2010 - link

    Why? For the lulz . . .
  • SofS - Tuesday, March 16, 2010 - link

    Firstly, a Linux overview is long overdue. From only an overall compatibility or a feature list to real CAD and maybe even some wine based tests (not to mention vdpau also). The wine tests can be helpful to the wine developers as I am sure yours are already to NVIDIA and ATI driver developers. On this regard, maybe even all open source developers could profit, from the X-server (Xorg-edgers) to interface (KDE, GNOME, XFCE, etc) and even driver (nouveau). Not to mention those nostalgic memories from 3ddesktop (nowadays resorting to COMPIZ).

    Secondly, GPGPU. My most pressing topic of interest as of now as I am myself researching in this area. Though you might need to dig through in order to find proper tests. Follow up comments with suggestions would be appreciated. My personal one would be folding@home GPU, as it is more disseminated and also runs over wine. By the way, it would be good to take double precision floating point (IEEE 754) into accont. From NVIDIA, as of now, only those 200 series GTX models offer such support and it is incomplete, it will be good to know how future ATI models and Fermi will compare.

    Thirdly, some CPU overhead tests. An example would be forceware running some game at x% CPU while catalyst at y% on the same game.

    Finally, testing those filters like AA and AF on and off are at least as important as how they scale at all levels (2x, 4x, 8x, 16x, transparency, et al). However, in order for those performance levels to make sense, the quality must be taken into account. Instead of using the number as an equivalent, use you quality analysis so that filters of nearly equivalent qualities are compared for performance.
  • marraco - Tuesday, March 16, 2010 - link

    I want an OpenCL benchmark.

    If it's a physic simulation, best.
  • PR3ACH3R - Tuesday, March 16, 2010 - link

    Games ? Games !?

    how about you test the current ATI DPC situation, DPC powerplay problems & DPC problems with software mode playback !

    http://www.avsforum.com/avs-vb/showthread.php?t=12...">http://www.avsforum.com/avs-vb/showthread.php?t=12...

    http://www.avsforum.com/avs-vb/showthread.php?t=12...">http://www.avsforum.com/avs-vb/showthread.php?t=12...

    How about you test what DXVA cards work with what formats,
    without BSODing like some currnet 5XXX series cards do.

    How about you test the horrible 2D performance with current ATI hardware.

    How about you make real articles like these people:

    http://translate.googleusercontent.com/translate_c...">http://translate.googleusercontent.com/...usg=ALkJ...

    http://translate.googleusercontent.com/translate_c...">http://translate.googleusercontent.com/...usg=ALkJ...
  • jive - Wednesday, March 17, 2010 - link

    chill out, the link you posted clearly states the solutions to the problems and for HTPC use the ATI seems to be the least bad choice anyway. No need to bash ATI on this front the 2D issue is already stated in previous posts.

    You're preaching to the choir here.

Log in

Don't have an account? Sign up now