Compute Performance

Moving on from our look at gaming performance, we have our customary look at compute performance. Since compute performance is by definition shader bound, the 7950 is at a bit of a disadvantage here compared to gaming performance. Whereas ROP performance scales with the core clock, shader performance is hit by both the reduction in the core clock and the disabled CU array.

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.

AMD’s greatly improved compute performance continues to shine here, though in the case of Civilization V it’s largely consumed by just closing the previously large gap between the GTX 500 series and the Radeon HD 6000 series. As a result the 7950 falls ever so short of the GTX 580, while the factory overclocked Sapphire and XFX cards give the 7950 enough of a push to come within 5% of the 7970.

Our next 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.

Under SmallLuxGPU the 7970 enjoyed a large lead over the GTX 580, and this continues with the 7950. Even though the 7950 is well behind the 7970—to the tune of 24%—it’s still 33% ahead of the GTX 580 and the lead only grows from there. Meanwhile the XFX and Sapphire cards can catch up to the 7970 somewhat, but as this is truly a shader-bound test, you can’t make up for the lack of shaders units on the 7950.

For our next benchmark we’re looking at AESEncryptDecrypt, an OpenCL AES encryption routine that AES encrypts/decrypts an 8K x 8K pixel square image file. The results of this benchmark are the average time to encrypt the image over a number of iterations of the AES cypher.

In spite of being a compute benchmark, AESEncryptDecrypt is not particularly sensitive to GPU performance, showcasing the impact that setup times can have. The 7950 trails the 7970 by 10%, and overclocking doesn’t change this much. Unfortunately for AMD NVIDIA is still the leader here, showing that AMD’s compute performance still has room to grow.

Finally, our last benchmark is once again looking at compute shader performance, this time through the Fluid simulation sample in the DirectX SDK. This program simulates the motion and interactions of a 16k particle fluid using a compute shader, with a choice of several different algorithms. In this case we’re using an (O)n^2 nearest neighbor method that is optimized by using shared memory to cache data.

With the compute shader fluid simulation we once again shift back into a compute task that’s much more shader-bound. The 7950 only reaches 80% of the performance of the 7970, once more proving the real impact of losing a CU array. This is still enough to handily surpass the GTX 580 however, with the 7950 taking a 15% lead.

Civilization V Power, Temperature, & Noise
Comments Locked

259 Comments

View All Comments

  • chizow - Sunday, February 5, 2012 - link

    Typical, resort to histrionics instead of facts and reasoning when arguing.

    Like I said, the value in the mid-range is there every single generation, not just with the 8800GT. There's a reason why AnandTech and many other sites proclaimed that launch "The Only Card that Matters" and since then any similar card that offers that same price:performance metric relative to the high end is affectionately referred to as "the next 8800GT" like the GTX 460/560 etc.
  • Galidou - Sunday, February 5, 2012 - link

    Still you won't take any arguments that will justify paying triple the price of a card for not even double the performance, or paying more for a card that offer less performance than a last gen x2 card NEVER.

    That gtx280 was really mispriced and Nvidia took people for dummies by pricing it so high for the performance you gained. The most performance you saw out of those was due to the first over 1gb memory in high resolution like 2560 which applied to not even .5% of the gaming population back then

    I don'T care about the titles, titles mean nothing to me that ''only card that matters'' doesn'T change the fact that paying triple the price for a video card for not even double the performance is tanking people for dummies.
  • Galidou - Sunday, February 5, 2012 - link

    We should be happy that Ati doesn't use such pricing technique because it competes against last gen refreshed video cards and justify to pay a higher % price than the % of performance you gain.
  • chizow - Sunday, February 5, 2012 - link

    Yeah they use far worst, at least this time.

    They offer linear scaling of price and performance instead of offering better performance at the same prices which you would expect from a next-gen chip.
  • chizow - Sunday, February 5, 2012 - link

    Why would I take any of your arguments when they are CLEARLY flawed.

    Its funny why don't you try using those arguments to justify the 7970's pricing? LMAO. Triple? Its not even DOUBLE the performance of previous mid-range cards.

    Or previous x2 cards? The 7970 isn't even in the same discussion as the 6990 or GTX 590.

    And of course, we've already covered how it falls completely short with regard to previous generation single-GPU flagship cards. Yet AMD somehow thinks it was worth a 10% increase in prices, something not even Nvidia has done since the 8800GTX.

    What's REALLY funny though is how you seem to think the GTX 280 was priced for dummies given it actually meets the standards for next-gen GPUs and deserved its price, yet see NOTHING wrong at all with the pricing of the 7970.

    Now what does that say about you and those defending it??? :(
  • Galidou - Sunday, February 5, 2012 - link

    It's not double the performance but not double the price either still 7950 performance of gtx580 for cheaper price, convinces me enough to say it isn't taking people for dummies...
  • chizow - Sunday, February 5, 2012 - link

    Yeah once again, selective application of flawed logic.

    Like I said, Tahiti meets none of the established standards you'd expect from a flagship part while the GTX 280.

    Yet you think the GTX 280 was priced for dummies and the 7970/7950's pricing is perfectly justified?

    Honestly that just makes you look dishonest.

    Or the exact type of customer AMD is looking to entice maybe! :D
  • Galidou - Sunday, February 5, 2012 - link

    LOL I'm not even looking at changing video card but I can say bad things about ATI but you'Re definitely 1: a hardcore Nvidia fanboy or 2: they pay you to say things about standard justifying stupid prices... lol standard, what an argument, next tie I'll buy a video card, I'll try to forget a little about perfromance and look at the standard...
  • chizow - Sunday, February 5, 2012 - link

    and galidou goes off the deepend folks.
  • Galidou - Sunday, February 5, 2012 - link

    mets the standard, now chizow is our standard expert, point is I look at numbers, not standard, triple the price not double the performance, I don'T care abbout the standard when I game with it on my computer.

    Oh wait my game stopped playing, it tells me, this video card doesn'T meet the standard for it'S price... OMG I should of cared more....

Log in

Don't have an account? Sign up now