• What
    is this?

    You've landed on the AMD Portal on AnandTech. This section is sponsored by AMD. It features a collection of all of our independent AMD content, as well as Tweets & News from AMD directly. AMD will also be running a couple of huge giveaways here so check back for those.

    PRESENTED BY

Image Quality: Anisotropic Filtering Tweaks & Tessellation Speed

Since the launch of Evergreen AMD has continued to tweak their anisotropic filtering quality. Evergreen introduced angle-independent filtering, and with the 6000 series AMD tweaked their AF algorithm to better handle high frequency textures. With Southern Islands that trend continues with another series of tweaks.

For Southern Islands AMD has changed the kernel weights of their anisotropic filtering mechanism in order to further reduce shimmering of high frequency textures. The algorithm itself remains unchanged and as does performance, but image quality is otherwise improved. Admittedly these AF changes seem to be targeting increasingly esoteric scenarios – we haven’t seen any real game where the shimmering matches the tunnel test – but we’ll gladly take any IQ improvements we can get.

Since AMD’s latest changes are focused on reducing shimmering in motion we’ve put together a short video of the 3D Center Filter Tester running the tunnel test with the 7970, the 6970, and GTX 580. The tunnel test makes the differences between the 7970 and 6970 readily apparent, and at this point both the 7970 and GTX 580 have similarly low levels of shimmering.


Video Download, H.264 (203MB)

While we’re on the subject of image quality, had you asked me two weeks ago what I was expecting with Southern Islands I would have put good money on new anti-aliasing modes. AMD and NVIDIA have traditionally kept parity with AA modes, with both implementing DX9 SSAA with the previous generation of GPUs, and AMD catching up to NVIDIA by implementing Enhanced Quality AA (their version of NVIDIA’s CSAA) with Cayman. Between Fermi and Cayman the only stark differences are that AMD offers their global faux-AA MLAA filter, while NVIDIA has support for true transparency and super sample anti-aliasing on DX10+ games.

Thus I had expected AMD to close the gap from their end with Southern Islands by implementing DX10+ versions of Adaptive AA and SSAA, but this has not come to pass. AMD has not implemented any new AA modes compared to Cayman, and as a result AAA and SSAA continue to only available in DX9 titles. And admittedly alpha-to-coverage support does diminish the need for these modes somewhat, but one only needs to fire up our favorite testing game, Crysis, to see the advantages these modes can bring even to DX10+ games. What’s more surprising is that it was AMD that brought AA IQ back to the forefront in the first place by officially adding SSAA, so to see them not continue that trend is surprising.

As a result for the time being there will continue to be an interesting division in image quality between AMD and NVIDIA. AMD still maintains an advantage with anisotropic filtering thanks to their angle-independent algorithm, but NVIDIA will have better anti-aliasing options in DX10+ games (ed: and Minecraft). It’s an unusual status quo that apparently will be maintained for quite some time to come.

Update: AMD has sent us a response in regard to our question about DX10+ SSAA

Basically the fact that most new game engines are moving to deferred rendering schemes (which are not directly compatible with hardware MSAA) has meant that a lot of attention is now being focused on shader-based AA techniques, like MLAA, FXAA, and many others. These techniques still tend to lag MSAA in terms of quality, but they can run very fast on modern hardware, and are improving continuously through rapid iteration.  We are continuing work in this area ourselves, and we should have some exciting developments to talk about in the near future.  But for now I would just say that there is a lot more we can still do to improve AA quality and performance using the hardware we already have.

Regarding AAA & SSAA, forcing these modes on in a general way for DX10+ games is problematic from a compatibility standpoint due to new API features that were not present in DX9.  The preferred solution would be to have games implement these features natively, and we are currently investigating some new ways to encourage this going forward.

Finally, while AMD may be taking a break when it comes to anti-aliasing they’re still hard at work on tessellation. As we noted when discussing the Tahiti/GCN architecture AMD’s primitive pipeline is still part of their traditional fixed function pipeline, and just as with Cayman they have two geometry engines that can process up to two triangles per clock. On paper at least Tahiti doesn’t significantly improve AMD’s geometry performance, but as it turns out there’s a great deal you can do to improve geometry performance without throwing more geometry hardware at the task.

For Southern Islands AMD has implemented several techniques to boost the efficiency of their geometry engines. A larger parameter cache is a big part of this, but AMD has also increased vertex re-use and off-chip buffering. As such while theoretical geometry throughput is unchanged outside of the clockspeed differences between 7970 and 6970, AMD will be making better use of the capabilities of their existing geometry pipeline.

By AMD’s numbers these enhancements combined with the higher clockspeed of the 7970 versus the 6970 give it anywhere between a 1.7x and 4x improvement in tessellation performance. In our own tests the improvements aren’t quite as great, but they’re still impressive. Going by the DX11DetailTessellation sample program the 7970 has better performance than the GTX 580 at both normal and high tessellation factors (and particularly at high tessellation factors), while under Unigine Heaven – a tessellation-heavy synthetic benchmark – the 7970 leads the GTX 580 by over 20%. Or compared to the 6970 the difference is even more stark, with the 7970 leading the 6970 by about 55% in both of these benchmarks.

Of course both of these benchmarks are synthetic and real world performance can (and will) differ, but it does prove that AMD’s improvements in tessellation efficiency really do matter. Even though the GTX 580 can push up to 8 triangles/clock, it looks like AMD can achieve similar-to-better tessellation performance in many situations with their Southern Islands geometry pipeline at only 2 triangles/clock.

Though with that said, we’re still waiting to see the “killer app” for tessellation in order to see just how much tessellation is actually necessary. Current games (even BF3) are DX10 games with tessellation added as an extra instead of being a fundamental part of the rendering pipeline. There are a wide range of games from BF3 to HAWX 2 using tessellation to greatly different degrees and none of them really answer the question of how much tessellation is actually necessary. Both AMD and NVIDIA have made tessellation performance a big part of their marketing pushes, so there’s a serious question over whether games will be able to utilize that much geometry performance, or if AMD and NVIDIA are in another synthetic numbers war.

Managing Idle Power: Introducing ZeroCore Power Drivers & ISV Relations
POST A COMMENT

291 Comments

View All Comments

  • mavere - Thursday, December 22, 2011 - link

    Seriously!

    With the prevalence of practically silent PSUs, efficient tower heatsinks, and large quiet fans, I cannot fathom why the noise floor is 37.9 dB.
    Reply
  • Finally - Thursday, December 22, 2011 - link

    As usual, AT is shooting straight for the brain-dam, I mean, ENTHUSIAST crowd feat. a non-mentioned power supply that should be well around 1000W in order to drive over-priced CPUs as well as quadruple GPU setups.
    If you find that horrendous they will offer you not to read this review, but their upcoming HTPC review where they will employ the same 1000W power supply...
    Reply
  • B3an - Thursday, December 22, 2011 - link

    *face palm*

    1: 1000+ Watt PSU's are normally more quiet if anything as they're better equipped to deal with higher power loads. When a system like this uses nowhere near the PSU's full power the fan often spins at a very low RPM. Some 1000+ PSU's will just shut the fan off completely when a system uses less than 30% of it's power.

    2: It's totally normal for a system to be around 40 dB without including the graphics cards. Two or 3 fans alone normally cause this much noise even if they're large low RPM fans. Then you have noise levels from surroundings which even in a "quiet" room are normally more than 15 dB.

    3: Grow some fucking brain cells kids.
    Reply
  • andymcca - Thursday, December 22, 2011 - link

    1) If you were a quiet computing enthusiast, you would know that the statement
    "1000+ Watt PSU's are normally more quiet if anything"
    is patently false. 1000W PSUs are necessarily less efficient at realistic loads (<600W at full load in single GPU systems). This is a trade-off of optimizing for efficiency at high wattages. There is no free lunch in power electronics. Lower efficiency yields more heat yields more noise, all else being equal. And I assure you that a high end silent/quiet PSU is designed for low air flow and uses components at least as high in quality as their higher wattage (non-silent/non-quiet) competitors. Since the PSU is not decribed (a problem which has been brought up many times in the past concerning AT reviews), who knows?

    2) 40dB is fairly loud if you are aiming for quiet operation. Ambient noise in a quiet room can be roughly 20dB (provided there is not a lot of ambient outdoor noise). 40dB is roughly the amplitude of conversation in a quiet room (non-whispered). A computer that hums as loud as I talk is pretty loud! I'm not sure if you opinion is informed by any empirical experience, but for precise comparison of different sources the floor should be at minimum 20dB below the sources in question.

    3) You have no idea what the parent's age or background is, but your comment #3 certainly implies something about your maturity.
    Reply
  • formulav8 - Tuesday, February 21, 2012 - link

    Seriously grow up. Your a nasty mouth as well. Reply
  • piroroadkill - Thursday, December 22, 2011 - link

    Haha, yeah.

    Still, I guess we have to leave that work to SPCR.
    Reply
  • Kjella - Thursday, December 22, 2011 - link

    High-end graphics cards are even noisier, so who cares? A 250W card won't be quiet no matter what. Using an overclocked Intel Core i7 3960X is obviously so the benchmarks won't be CPU limited, not to make a quiet PC. Reply
  • Ryan Smith - Thursday, December 22, 2011 - link

    Our testing methodology only has us inches from the case (an open case I should add), hence the noise from our H100 closed loop radiator makes itself known. In any case these numbers aren't meant to be absolutes, we only use them on a relative basis. Reply
  • MadMan007 - Thursday, December 22, 2011 - link

    [AES chart] on page 7? Reply
  • MadMan007 - Thursday, December 22, 2011 - link

    More stuff missing on page 9:

    [AF filter test image] [download table]
    Reply

Log in

Don't have an account? Sign up now