Compute Performance

As always we'll start with our DirectCompute game example, 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.  While DirectCompute is used in many games, this is one of the only games with a benchmark that can isolate the use of DirectCompute and its resulting performance.

Compute: Civilization V

As our Civilization V compute benchmark is just that, a compute benchmark, so our results aren’t too surprising. This is one of the few compute tests NVIDIA does well at, so the GTX 650 Ti Boost is close to both Radeon cards, and not all that far behind the GTX 660 either.

Our next benchmark is LuxMark2.0, the official benchmark of SmallLuxGPU 2.0. SmallLuxGPU is an OpenCL accelerated ray tracer that is part of the larger LuxRender suite. Ray tracing has become a stronghold for GPUs in recent years as ray tracing maps well to GPU pipelines, allowing artists to render scenes much more quickly than with CPUs alone.

Compute: LuxMark 2.0

Moving on to LuxMark, we quite frankly transition into a more normal compute benchmark pattern for NVIDIA, which sees Kepler flopping. The GTX 650 Ti Boost can’t get even remotely close to a 7770, let alone the 7850. On the NVIDIA side it doesn’t help that since this is a compute benchmark the GTX 650 Ti Boost gains fairly little over the GTX 650 Ti.

Our 3rd benchmark set comes from CLBenchmark 1.1. CLBenchmark contains a number of subtests; we’re focusing on the most practical of them, the computer vision test and the fluid simulation test. The former being a useful proxy for computer imaging tasks where systems are required to parse images and identify features (e.g. humans), while fluid simulations are common in professional graphics work and games alike.

Compute: CLBenchmark 1.1 Computer Vision

Compute: CLBenchmark 1.1 Fluid Simulation

CLBenchmark is much the same as LuxMark, with NVIDIA cards bringing up the rear. The fluid simulation ends up being the more painful of the two benchmarks for the GTX 650 Ti Boost, clocking in at less than 1/3rd the performance of the 7850.

Moving on, our 4th compute benchmark is FAHBench, the official Folding @ Home benchmark. Folding @ Home is the popular Stanford-backed research and distributed computing initiative that has work distributed to millions of volunteer computers over the internet, each of which is responsible for a tiny slice of a protein folding simulation. FAHBench can test both single precision and double precision floating point performance, with single precision being the most useful metric for most consumer cards due to their low double precision performance. Each precision has two modes, explicit and implicit, the difference being whether water atoms are included in the simulation, which adds quite a bit of work and overhead. This is another OpenCL test, as Folding @ Home is moving exclusively OpenCL this year with FAHCore 17.

Compute: Folding @ Home: Explicit, Single Precision

Compute: Folding @ Home: Implicit, Single Precision

NVIDIA still struggles at compute with FAHBench – the move to OpenCL isn’t doing them any favors – but it’s not the blowout that was our last two benchmarks. Interestingly explicit favors NVIDIA more than implicit, which may mean NVIDIA is handling the overhead better than AMD is. Still, any Folding @ Home users will be far better served by AMD than NVIIDA here.

Our 5th compute benchmark is Sony Vegas Pro 12, an OpenGL and OpenCL video editing and authoring package. Vegas can use GPUs in a few different ways, the primary uses being to accelerate the video effects and compositing process itself, and in the video encoding step. With video encoding being increasingly offloaded to dedicated DSPs these days we’re focusing on the editing and compositing process, rendering to a low CPU overhead format (XDCAM EX). This specific test comes from Sony, and measures how long it takes to render a video.

Compute: Sony Vegas Pro 12 Video Render

Vegas is another OpenCL benchmark, and another benchmark NVIDIA brings up the rear with. Certainly the additional compute performance of the GTX 650 Ti Boost over the GTX 650 Ti is helping NVIDIA here, but it can’t make up for a gap of over 30 seconds.

Wrapping things up, our final compute benchmark is an in-house project developed by our very own Dr. Ian Cutress. SystemCompute is our first C++ AMP benchmark, utilizing Microsoft’s simple C++ extensions to allow the easy use of GPU computing in C++ programs. SystemCompute in turn is a collection of benchmarks for several different fundamental compute algorithms, as described in this previous article, with the final score represented in points. DirectCompute is the compute backend for C++ AMP on Windows, so this forms our other DirectCompute test.

Compute: SystemCompute v0.5.7.2 C++ AMP Benchmark

SystemCompute mixes things up a bit with its multiple sub-benchmarks, but it still doesn’t change the fact that Kepler and GTX 650 Ti Boost just don’t do that well in most compute scenarios. 68K points is enough to tie the 6870 of all things, itself not a particular good compute card. Otherwise the bar is set by AMD at over 100K points.

Civilization V Synthetics
POST A COMMENT

78 Comments

View All Comments

  • Eugene86 - Tuesday, March 26, 2013 - link

    Overclocking is not a guarantee, apart from the fact that it shortens the lifespan of the card itself, all it does is increase power consumption (which negates your power consumption argument) and it also wastes your time by having to have to tinker with drivers and overclocking speeds to make sure the card doesn't artifact and such. Considering the subpar quality of AMD drivers, this is something that no one wants to waste their time on except for fans of AMD. Reply
  • stickmansam - Tuesday, March 26, 2013 - link

    Have you tried to OC the 7850 yourself? I hit 1ghz without any voltage or power tweaks
    It is way under clocked at 860mhz.

    Haven't run any power draw comparisons yet though myself
    Reply
  • cyrusfox - Tuesday, March 26, 2013 - link

    So true, 7850 hits 1050 MHz easy breezy. It was a great card and it has held its value, I originally got it for $150 second hand when it first came out(forum shopping). I'm surprised to see it still being valued higher than that.

    From my personal experience, the 7850 is a guaranteed 1 GHz card. But a lot of people don't OC so they miss out on the potential. Nvidia does have its positives and the boost seems like a good card to fill the holes in the market.
    Reply
  • anubis44 - Thursday, March 28, 2013 - link

    I hear what you're saying -- in the past, people would say that you could overclock a model of card to some ridiculous speed that only 5% of the cards of that model could reach with stability, and then say that the entire model of card performed that way. A couple of times, I was tricked by this kind of nonsense, and regretted swallowing it from idiots on a chat forum.

    But in the case of the 7850, this is the real deal. I kid you not, there is probably not a single one that won't go up to 1050MHz with just the AMD video driver and a couple of fisher-price sliders and not even any additional voltage. I defy you to find somebody who has tried this, and been unable to clock it up to at least the 1050MHz max in the video driver, without any extra volts. It's the best kept secret in the video card market right now, almost like the Celeron 300a back in the day, that would clock from 300MHz to 450MHz with just a multiplier adjustment in the bios. It's like AMD just left 40% more performance on the table with the 7850. I had two of them myself for about a week, and it was ridiculous how much faster they would go if you just bothered to take 2 minutes to bump them up from 860MHz to 1100MHz-1150MHz. I can picture AMD getting slightly irritated at nVidia for trying to reach the 7850's default speed with this piddly GTX650 Ti Boost, and turning around and releasing a GHz edition bios for all 7850 owners to just flash their cards to 1050MHz, kinda of like they did for the 7950.
    Reply
  • k2_8191 - Tuesday, March 26, 2013 - link

    IMO The new card is good deal if I consider only gaming performance per price.
    However, as a distributed computing junkie, I would still recommend the participants to choose some RADEON HD 7k cards for double-precision workunits (HCC in WCG in particular).
    Reply
  • Bob Todd - Tuesday, March 26, 2013 - link

    One simple and obvious answer, Never Settle Reloaded. The 7850 is faster in most games, destroys it compute, and you get Tomb Raider and BioShock Infinite for free. Even if you have no interest in either game, you can easily get ~$50 (after fees) for the codes on eBay. Nice 2GB 7850s have been on sale regularly for $170. That means you can get a faster 7850 2GB for ~$120. So does $150 for a 1GB 650 Ti Boost sound better than $120 for a 2GB 7850? 1GB 7850s have hit $150. Bottom line is that it's great time to be a budget gamer. The 650 Ti Boost, 7790, and 7850 are all good cards in the pricing sweet spot for a lot of people. Never Settle Reloaded puts things squarely in AMDs hands from a "value" perspective though. Whether or not you want the games, they are quick sells, and even the 7790 would really be in the ~$120 range if you wanted to sell BioShock Infinite. Reply
  • Parhel - Tuesday, March 26, 2013 - link

    Thanks! Based on your post, I did a bit of shopping, and found an MSI 7870 @ Newegg. $220 list price minus a $25 rebate, ~$50 on reselling the bundle, and ~$30 on reselling my GTX 460, and I should be at about $115 for a 7870. Not too bad considering they were ~$350 at release. Reply
  • Bob Todd - Tuesday, March 26, 2013 - link

    You just reminded me that I need to decide if I'm selling my GTX 460. I looked on eBay a few weeks ago and the same 1GB model sold for ~$60. Part of me thinks it isn't really worth it and I should just keep it as a backup or a hand-me-down upgrade for a family member. But the smarter part of me knows it will probably just sit in its box in my closet for years. Reply
  • just4U - Wednesday, March 27, 2013 - link

    I think you can get a bit more for the 460.. $60 is a fair price for it.. imo. Also "HIS" seems to offer the best deals on the 7870 (when they come up for sale..) Not sure if you can really get $50 for the bundle tho.. hah can always try! Reply
  • Bob Todd - Wednesday, March 27, 2013 - link

    YMMV of course. However I've sold two for that much, with the listings lasting less than 12 hours. And most people are happy to just get the code via email so you might not even have to ship anything. Reply

Log in

Don't have an account? Sign up now