NVIDIA's GeForce GTX 580: Fermi Refined
by Ryan Smith on November 9, 2010 9:00 AM ESTCompute and Tessellation
Moving on from our look at gaming performance, we have our customary look at compute performance, bundled with a look at theoretical tessellation performance. Unlike our gaming benchmarks where NVIDIA’s architectural enhancements could have an impact, everything here should be dictated by the core clock and SMs, with shader and polymorph engine counts defining most of these tests.
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.
We previously discovered that NVIDIA did rather well in this test, so it shouldn’t come as a surprise that the GTX 580 does even better. Even without the benefits of architectural improvements, the GTX 580 still ends up pulling ahead of the GTX 480 by 15%. The GTX 580 also does well against the 5970 here, which does see a boost from CrossFire but ultimately falls short, showcasing why multi-GPU cards can be inconsistent at times.
Our second compute benchmark is Cyberlink’s MediaEspresso 6, the latest version of their GPU-accelerated video encoding suite. MediaEspresso 6 doesn’t currently utilize a common API, and instead has codepaths for both AMD’s APP (née Stream) and NVIDIA’s CUDA APIs, which gives us a chance to test each API with a common program bridging them. As we’ll see this doesn’t necessarily mean that MediaEspresso behaves similarly on both AMD and NVIDIA GPUs, but for MediaEspresso users it is what it is.
We throw MediaEspresso 6 in largely to showcase that not everything that’s GPU accelerated is GPU-bound, as ME6 showcases this nicely. Once we move away from sub-$150 GPUs, APIs and architecture become much more important than raw speed. The 580 is unable to differentiate itself from the 480 as a result.
Our third 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 is rather straightforward in its requirements: compute and lots of it. The GTX 580 attains most of its theoretical performance improvement here, coming in at a bit over 15% over the GTX 480. It does get bested by a couple of AMD’s GPUs however, a showcase of where AMD’s theoretical performance advantage in compute isn’t so theoretical.
Our final compute benchmark is a Folding @ Home benchmark. Given NVIDIA’s focus on compute for Fermi and in particular GF110 and GF100, cards such as the GTX 580 can be particularly interesting for distributed computing enthusiasts, who are usually looking for the fastest card in the coolest package. This benchmark is from the original GTX 480 launch, so this is likely the last time we’ll use it.
If I said the GTX 580 was 15% faster, would anyone be shocked? So long as we’re not CPU bound it seems, the GTX 580 is 15% faster through all of our compute benchmarks. This coupled with the GTX 580’s cooler/quieter design should make the card a very big deal for distributed computing enthusiasts.
At the other end of the spectrum from GPU computing performance is GPU tessellation performance, used exclusively for graphical purposes. Here we’re interesting in things from a theoretical architectural perspective, using the Unigine Heaven benchmark and Microsoft’s DirectX 11 Detail Tessellation sample program to measure the tessellation performance of a few of our cards.
NVIDIA likes to heavily promote their tessellation performance advantage over AMD’s Cypress and Barts architectures, as it’s by far the single biggest difference between them and AMD. Not surprisingly the GTX 400/500 series does well here, and between those cards the GTX 580 enjoys a 15% advantage in the DX11 tessellation sample, while Heaven is a bit higher at 18% since Heaven is a full engine that can take advantage of the architectural improvements in GF110.
Seeing as how NVIDIA and AMD are still fighting about the importance of tessellation in both the company of developers and the public, these numbers shouldn’t be used as long range guidance. NVIDIA clearly has an advantage – getting developers to use additional tessellation in a meaningful manner is another matter entirely.
Facebook
Twitter
RSS
160 Comments
View All Comments
deeps6x - Tuesday, November 9, 2010 - link
Really, who has to have this card when for less money, you can get better results with a pair of 460s in SLI or a pair of 6850s in CF (even better than 460s in almost all cases) to give you better numbers than this card.Take the extra $100+ and bump your CPU up a speed bin or two.
Alberto8793 - Tuesday, November 9, 2010 - link
I will give Nvidia some props they have a very fast card, but interms of performance per dollar, AMD beats them out, go to newegg and click on the 5970 its 499$, go and check the 580, a cool 560$ now return to the anandtech review and check out the benchmarks, they speak for themselves, and finnally load and idle consumption of a two gpu card are lower than a single gpu?!?!?!?!? OUTRAGEOUS!slick121 - Wednesday, November 10, 2010 - link
I mean this is the "enough said!" post. So true, totally agree!ClagMaster - Tuesday, November 9, 2010 - link
Looks like Nvidia has done their homework optimize the GF100 after business pressures to release the GTX-480.Nvidia did a competent makeover of the GF100 and GTX-480 to reduce power, heat and noise.
The GF-110/GTX-580 offers about 10% more performance for the Watts, which are still too high. for my liking.
AnnonymousCoward - Wednesday, November 10, 2010 - link
I think it's more like 30% more performance per watt. Crysis power is 389W vs 421W and averages 15% more fps.the_elvino - Wednesday, November 10, 2010 - link
I would take Anand's power consumtion measurements of 421W for the GTX 480 and 389W for GTX 590 with a few pinches of salt.HardOCP compared the power draw in SLI and recorded following numbers for the entire system under full load:
GTX 480 SLI: 715W
GTX 590 SLI: 700W
Difference per card between 480 and 590: 7,5W!
The 590 is noticeably cooler though.
http://hardocp.com/article/2010/11/09/nvidia_gefor...
7Enigma - Wednesday, November 10, 2010 - link
Why would HardOCP's numbers be any more accurate than Anand's? One is comparing SLI and one is single card. We have no clue the differences in system configuration (individual variances of parts not the part name), not to mention the inherent variability of the cards themselves.I've said it in the 6850 article and before but power consumption numbers (and thus noise/heat) have to be taken with a grain of salt because there is no independent unbiased source giving these cards for review. A really good or poor cherry-picked sample can completely change the impression of a cards efficiency. And since we know NVIDIA ships cards with varying load voltages it is very easy to imagine a situation where one card would be similar in power draw to the 480 while another would show lower thermals.
Anand/Ryan I've said it before and I'll say it again, you have the chance to be a pioneer in the review field by getting independent sources of cards slightly after release. Yes your original review would have to exclude this (or have a caveat that this is what the manufacturer supplied you with).
Decide on a good number (3-5 shouldn't be unreasonable), and purchase these from etailers/locally anonymously. Test them for power/noise/temp and then either sell them on Ebay at slightly reduced cost or even increased price? (Anandtech approved!...touched by Ryan himself!....whatever....you've got a marketing department) :), and/or take one/all and give them away to your loyal readers like the other contests.
Then have a followup article with the comparisons between the "real" cards and the cherry-picked ones from NVIDIA/AMD. Just think how if the manufacturers know that you might call them out on a cherry-picked sample in a follow-up review, they might just try to avoid the bad press and ship one more closely in line with the product line.
Goblerone - Tuesday, November 9, 2010 - link
I'm confused here. In your review of ATI's new 6870, you were comparing against factory overclocked NVIDIA cards but here you are using reference ATI cards.What gives?
AnandThenMan - Wednesday, November 10, 2010 - link
Why even bother asking. It's clear that Anandtech is not about fairness in reviews anymore. Nvidia got special treatment, plain and simple.Just wait until we see Anandtech preview AMD's Brazos with some benches, you will see even more bizarre and head scratching benchmarks. Wait and see.
Sihastru - Wednesday, November 10, 2010 - link
Lol, give it a rest man... and while you're at it tells us about some similarly priced factory overclocked cards that AMD has on the stores shelves and could be used at the time the review was conducted. Relevent models only please, that have the same performance as the GTX580.