Compute

Moving on from our look at gaming performance, we have our customary look at compute performance.

Our first compute benchmark comes from Civilization V, which uses DirectCompute to decompress textures on the fly. Civ5 includes a sub-benchmark that exclusively tests the speed of their texture decompression algorithm by repeatedly decompressing the textures required for one of the game’s leader scenes.

In the game world Civ5 benefits significantly from SLI and CrossFire. For our texture compression test however AFR is more a liability than a benefit. This doesn’t impact the game in any meaningful manner, but it’s an example of how SLI/CF aren’t always the right tool for the job. Unfortunately for both parties, with as few compute applications as there are today, almost none of them benefit from SLI/CF.

Our second GPU compute benchmark is SmallLuxGPU, the GPU ray tracing branch of the open source LuxRender renderer. While it’s still in beta, SmallLuxGPU recently hit a milestone by implementing a complete ray tracing engine in OpenCL, allowing them to fully offload the process to the GPU. It’s this ray tracing engine we’re testing.

SmallLuxGPU only currently supports ray tracing with one GPU, so all of our results are effectively proxies for what would be if the GTX 590 only had one GPU. Not surprisingly overclocks do wonders here, and NVIDIA’s strong compute architecture gives them an easy win. SLI/CF performance will become more important here when we upgrade to LuxMark for our next iteration of our benchmark suite, as LuxMark can handle multiple OpenCL drivers.

Our final compute benchmark is a Folding @ Home benchmark. Given NVIDIA’s focus on compute for Fermi, cards such as the GTX 590 can be particularly interesting for distributed computing enthusiasts, as two GPUs should be able to quickly retire work units.

Folding@Home doesn’t directly benefit from CF/SLI at all. However by dispatching one WU to each GPU it’s possible to double effective performance. With that taken into account the GTX 590 is quite an effective cruncher, particularly when we start looking at overclocking.

Wolfenstein Power, Temperature, & Noise
POST A COMMENT

123 Comments

View All Comments

  • Ryan Smith - Thursday, March 24, 2011 - link

    One way or another we will be including multi-monitor stuff. The problem right now is getting ahold of a set of matching monitors, which will take some time to resolve. Reply
  • fausto412 - Thursday, March 24, 2011 - link

    also would be nice to test 1680x1050 on at least a couple of demanding games. illustrate to people who have 22" screens that these cards are a waste of money at their resolution. Reply
  • bigboxes - Thursday, March 24, 2011 - link

    It has been a waste for that low resolution since two generations ago. But you knew that. Troll... Reply
  • tynopik - Thursday, March 24, 2011 - link

    matching monitors might matter for image quality or something, but for straight benchmarking, who cares?

    surely you have 3 monitors capable of 1920x1080

    it's not like the card cares if one is 20" and another is 24"
    Reply
  • 7Enigma - Thursday, March 24, 2011 - link

    I don't understand this either. There is no need for anything fancy, heck you don't even need to have them actually outputting anything, just fool the drivers into THINKING they are driving multiple monitors! Reply
  • DanNeely - Thursday, March 24, 2011 - link

    I don't entirely agree. While it doesn't matter much for simple average FPS benches like Anandtech is currently doing, they fall well short of the maximum playable settings testing done by sites like HardOCP. Reply
  • strikeback03 - Thursday, March 24, 2011 - link

    Remember, the AT editors are spread all over. So while between them they certainly have at least 3 1920x1080/1200 monitors, Ryan (doing the testing) probably doesn't.

    Plus with different monitors wouldn't response times possibly be different? I'd imagine that would be odd in gaming.
    Reply
  • tynopik - Thursday, March 24, 2011 - link

    > Remember, the AT editors are spread all over. So while between them they certainly have at least 3 1920x1080/1200 monitors, Ryan (doing the testing) probably doesn't.

    This has been a need for a while, and it's not like this review was completely unexpected, so not sure why they don't have a multi-monitor setup yet

    > Plus with different monitors wouldn't response times possibly be different? I'd imagine that would be odd in gaming.

    Well that's sort of the point, they wouldn't actually be gaming, so who cares?
    Reply
  • Martin Schou - Thursday, March 24, 2011 - link

    I would have thought that the marketing departments of companies like Asus, Benq, Dell, Eizo, Fujitzu, HP, LaCie, LG, NEC, Philips, Samsung and ViewSonic would cream their pants at what is really very cheap PR.

    Supply sets of 3 or 5 1920x1080/1920x1200 displays and 3 or 5 2560x1440/2560x1600 displays in exchange for at least a full year's advertisement on a prominent tech website.

    If we use Dell as an example, they could supply a set of five U2211H and three U3011 monitors for a total cost of less than 5,900 USD per set. The 5,900 USD is what us regular people would have to pay, but in a marketing campaign it's really just a blip on the radar.

    Now, excuse me while I go dream of a setup that could pull games at 9,600x1080/5,400x1920 or 7,680x1600/4,800x2560 :D
    Reply
  • Ryan Smith - Friday, March 25, 2011 - link

    I'd just like to note that advertising is handled separately from editorial content. The two are completely compartmentalized so that ad buyers can't influence editorial control. Conversely as an editor I can't sell ad space. Reply

Log in

Don't have an account? Sign up now