NVIDIA GeForce GTX 690 Review: Ultra Expensive, Ultra Rare, Ultra Fast
by Ryan Smith on May 3, 2012 9:00 AM ESTCompute Performance
For our look at compute performance this is going to be a brief look. Our OpenGL AES and DirectCompute Fluid Simulation benchmarks simply don’t scale with multiple GPUs, so we’ll skip though (though the data is still available in Bench).
Our first compute benchmark comes from Civilization V, which uses DirectCompute to decompress textures on the fly. Civ V 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. Note that this is a DX11 DirectCompute benchmark.
Given the nature of the benchmark, it’s not surprising that we see a performance regression here with some setups. The nature of this benchmark is that it doesn’t split across multiple GPUs well, though that doesn’t stop AMD and NVIDIA from tying. This doesn’t impact real game performance as we’ve seen, but it’s a good reminder of the potential pitfalls of multi-GPU configurations. Though AMD does deserve some credit here for gaining on their single GPU performance, pushing their lead even higher.
Our other compute benchmark is SmallLuxGPU, the GPU ray tracing branch of the open source LuxRender renderer. We’re now using a development build from the version 2.0 branch, and we’ve moved on to a more complex scene that hopefully will provide a greater challenge to our GPUs.
Unlike the Civ V compute benchmark, SLG scales very well with multiple GPUs, nearly doubling in performance. Unfortunately for NVIDIA GK104 shows its colors here as a compute-weak GPU, and even with two of them we’re nowhere close to one 7970, let alone the monster that is two. If you’re looking at doing serious GPGPU compute work, you should be looking at Fermi, Tahiti, or the future Big Kepler.
Facebook
Twitter
RSS
200 Comments
View All Comments
CeriseCogburn - Sunday, May 6, 2012 - link
It's a very pertinent topic because despite the protestations the vast majority of readers are likely to have a 1080p monitor, and apparently they all have amnesia as well since this came up in a very, very recent article - one of the last few on the video cards - where it was even pointed out by Anand himself that one should extrapolate the fps data with the 11% pixel difference, something every last one of you either never read or completely forgot or conveniently omitted.Unfortunately Anand isn't correct, period. He should know it, but of course has to have an answer for the complainers - so he gave one.
What should be clear, even by just looking at the 1920X benches here is that below that resolutin and above it don't always jibe with it - and personally I've noticed nVidia has issues with 1600X1200.
So, all the hundreds of readers with 1080p monitors can thank me deep down in their hearts, as they now know Nvidia is supreme at that resolution, by 17%+ on average, more so than at 1200p, and thus "saving money" on an amd card using this sites 1920 1200p stats is for many, incorrect.
Makaveli - Thursday, May 3, 2012 - link
no one is mad or raging I think most are just amused at your stupidity! and sad attempt at trolling.Trololol
CeriseCogburn - Sunday, May 6, 2012 - link
Yet my point is absolutely prescient and helpful to anyone whom isn't a raging fanboy fool.Doesn't appear you've risen above that category.
InsaneScientist - Sunday, May 6, 2012 - link
And yet they specifically called out the fact that the patch broke performance on nVidia cards, went out of their way to state what performance was like before the patch (which is clearly better than any of the other cards), and finally stated that they're pretty sure that the game is at fault, not nVidia or their drivers...Yeah, they really must have it out for nVidia... *sigh*
CeriseCogburn - Sunday, May 6, 2012 - link
Except in the 680 tests all you fools ignored S2TW, which I had to repeatedly point out was the hardest game in the bunch, not Crysis or M2033 - over and over again I had to tell all the fools blabbering, and now suddenly the game is "broken". ROFL - it's beyond amazing.CeriseCogburn - Sunday, May 6, 2012 - link
Oh look it's not broken, TPU can handle it with nearly 100% single card SLI scaling with the 690http://www.techpowerup.com/reviews/NVIDIA/GeForce_...
Gee I guess it was "harder" here.
blackened23 - Thursday, May 3, 2012 - link
you appear to be using an old version of Batman without the latest patch.blackened23 - Thursday, May 3, 2012 - link
http://www.computerandvideogames.com/340746/batman...You are using an outdated version, the newest version released in March enhances both SLI and CF performance.
Ryan Smith - Thursday, May 3, 2012 - link
We're on the latest version. I triple checked.blackened23 - Thursday, May 3, 2012 - link
Are you using the Steam version? Your results differ from that of HardOCP, hardwareheaven, and hardwarecanucks. They get scaling you don't. Your version should be dated March 2012, thats when the patch was released.