Ashes Of The Singularity Beta: Async Compute, Multi-Adapter & Power

DirectX 12 has been available since Windows 10, but there aren't any games for it yet, so we're using the Ashes of the Singularity beta to examine DX12 performance.

We've benchmarked unfinished games several times in our reviews, most recently in our Radeon R9 Nano launch article. Back then, we found some driver issues and, even worse, hardware problems on Nvidia graphics cards. It didn't take that company long to respond with new software that not only helped overcome those technical shortcomings, but even turned the performance story around, enabling a small lead over the competition.

This time around was a bit different. Since AMD provided us with an optimized launch driver, we asked Nvidia if it had one for us as well. Company reps didn't seem too pleased with the question and pointed us to its current WHQL-certified driver, version 361.91. That was all Nvidia would say on the matter.

As a game, Ashes of the Singularity is really the perfect showcase for DirectX 12. It features many small AI-controlled entities moving across a large area. This is both a pro and a con. It does show off the API's benefits nicely. But we also have to recognize that those gains will probably be less pronounced in titles that don't exploit DirectX 12's new features as heavily.

The capability making the headlines lately is called Asynchronous Shading/Compute. It allows the parallel and asynchronous (meaning completely order-independent) execution of graphical tasks (shading) and calculations (compute). If correctly implemented in hardware and software, the technology can cut latency down by a massive amount, which in turn results in higher performance.

Let's review a diagram of how tasks are handled in DirectX 11. They are executed one by one in a fixed order. This order can't be changed, and the only way to achieve efficiency is to split up the queue and keep it relatively short.

In DirectX 12, the queue can be split so that tasks are completed at the same time and somewhat offset from each other. This works fairly well for the benchmark we're using today, so we'll actually see its potential in practice. Just remember, though, how much of a difference this new feature makes is going to depend on how much the tasks in a particular game benefit from being executed in parallel, which is to say asynchronously. There won't be any benefit if tasks are dependent on other tasks' results, or if the overhead involved in managing all the tasks is higher than the gains from the asynchronous execution.

Several revisions of AMD's GCN architecture include provisions for this functionality, so its Tonga, Hawaii and Fiji GPUs fare the best with asynchronous shading/compute. Nvidia's Kepler and Maxwell architectures are having a much harder time. The company is trying to compensate with software-based solutions instead.

The following graph clearly shows that Nvidia isn't there yet. After activating the new DirectX 12 feature in the .ini file, AMD's performance increases markedly, whereas Nvidia's actually gets slightly worse. The option to switch this setting on and off convinced us to leave DirectX 11 alone completely and focus all of our efforts on DirectX 12 and asynchronous compute. To keep things fair, we're testing all graphics cards with the setting that works best for them.

MORE: Best Graphics Cards
MORE: All Graphics Content

Test System

We're using the same test system that we've been using for quite a while. We didn't make any changes to it:

Technical Specifications
Test System
Intel Core i7-5930K at 4.2GHz
Alphacool Water Cooler (NexXxos CPU Cooler, VPP655 Pump, Phobya Balancer, 24cm Radiator)
Crucial Ballistix Sport, 4x 4GB DDR4-2400
MSI X99S XPower AC
1x Crucial MX200, 500GB SSD (System)
1x Corsair Force LS 960GB SSD (Applications, Data)
be quiet Dark Power Pro, 850W PSU
Windows 10 Pro (All Updates)
Drivers
AMD: Radeon Software 15.301 B35 (Press Beta Driver, February 2016)
Nvidia: ForceWare 361.91 WHQL
Gaming
Benchmarks
Ashes of the Singularity Beta 2 (Press)

Since the benchmark provides us with a very interesting log file that contains a lot of details (and not just the overall results), we updated our interpreter to take full advantage of it by looking at important things like frame times.

Create a new thread in the US Reviews comments forum about this subject
This thread is closed for comments
116 comments
    Your comment
    Top Comments
  • FormatC
    Quote:
    An AMD sponsored title
    Really? Sponsoring and knowledge sharing are two pairs of shows. Nvidia was invited too. ;)

    Quote:
    Async Compute will quickly start to take the same road taken by Mantle

    Sure? The design of the most current titles and engines was started long time before Microsoft started with DirectX 12. You can find the DirectX 12 render path in first steps now in a lot of common engines and I'm sure that PhysX will faster die than DirectX12. Mantle was the key feature to wake up MS, not more. And: it's async compute AND shading :)
    16
  • ohim
    Quote:
    An AMD sponsored title that shows off the one and only part of DirectX 12 where AMD cards have an advantage. The key statement is: "But where are the games that take advantage of the technology?" Without that, Async Compute will quickly start to take the same road taken by Mantle, remember that great technology?

    Instead of making random assumptions about the future of DX12 and Async shaders you should first be mad at Nvidia for stating they have full DX12 cards and that`s not the case, and the fact that Nvidia is trying hard to fix this issues trough software tells a lot.

    PS: it`s so funny to see the 980ti being beaten by 390x :)
    10
  • chimera201
    Such research. Much detail. Nice benchmark analysis TH.
    10
  • Other Comments
  • freak777power
    In other words DX12 is business gimmick which doesn't translate to squat in real game scenario and I am glad I stayed on Windows 7...running crossfire R9 390x.
    -19
  • FormatC
    Especially Hawaii / Grenada can benefit from asynchronous shading / compute (and your energy supplier). :)
    0
  • 17seconds
    An AMD sponsored title that shows off the one and only part of DirectX 12 where AMD cards have an advantage. The key statement is: "But where are the games that take advantage of the technology?" Without that, Async Compute will quickly start to take the same road taken by Mantle, remember that great technology?
    -22
  • FormatC
    Quote:
    An AMD sponsored title
    Really? Sponsoring and knowledge sharing are two pairs of shows. Nvidia was invited too. ;)

    Quote:
    Async Compute will quickly start to take the same road taken by Mantle

    Sure? The design of the most current titles and engines was started long time before Microsoft started with DirectX 12. You can find the DirectX 12 render path in first steps now in a lot of common engines and I'm sure that PhysX will faster die than DirectX12. Mantle was the key feature to wake up MS, not more. And: it's async compute AND shading :)
    16
  • turkey3_scratch
    Well, there's no denying that for this game the 390X sure is great performing.
    9
  • James Mason
    Quote:
    Quote:
    An AMD sponsored title
    Really? Sponsoring and knowledge sharing are two pairs of shows. Nvidia was invited too. ;)

    Quote:
    Async Compute will quickly start to take the same road taken by Mantle

    Sure? The design of the most current titles and engines was started long time before Microsoft started with DirectX 12. You can find the DirectX 12 render path in first steps now in a lot of common engines and I'm sure that PhysX will faster die than DirectX12. Mantle was the key feature to wake up MS, not more. And: it's async compute AND shading :)


    Geez, Phsyx has been around for so long now and usually only the fanciest of games try and make use of it. It seems pretty well adopted, but it's just that not all games really need to add an extra layer of physics processing "just for the lulz."
    -3
  • Wisecracker
    Thanks for the effort, THG! Lotsa work in here.

    What jumps out at me is how the GCN Async Compute frame output for the R9 380X/390X barely moves from 1080p to 1440p ---- despite 75% more frames. That's sumthin' right there.

    It will be interesting to see how Pascal responds ---- and how Polaris might *up* AMD's GPU compute.

    Neat stuff on the CPU, too. It would be interesting to see how i5 ---> i7 hyperthreads react, and how the FX 8-cores (and 6-cores) handle the increased emphasis on parallelization.

    You guys don't have anything better to do .... right? :)
    8
  • freak777power
    For someone who runs Crossfire R9 390x (three cards) DX12 makes no difference in term of performance. For all BS Windows 10 brings not worth *downgrading to considering that lot of games under Windows 10 are simply broken or run like garbage where no issue under Windows 7.
    -6
  • ohim
    Quote:
    An AMD sponsored title that shows off the one and only part of DirectX 12 where AMD cards have an advantage. The key statement is: "But where are the games that take advantage of the technology?" Without that, Async Compute will quickly start to take the same road taken by Mantle, remember that great technology?

    Instead of making random assumptions about the future of DX12 and Async shaders you should first be mad at Nvidia for stating they have full DX12 cards and that`s not the case, and the fact that Nvidia is trying hard to fix this issues trough software tells a lot.

    PS: it`s so funny to see the 980ti being beaten by 390x :)
    10
  • cptnjarhead
    Quote:
    For someone who runs Crossfire R9 390x (three cards) DX12 makes no difference in term of performance. For all BS Windows 10 brings not worth *downgrading to considering that lot of games under Windows 10 are simply broken or run like garbage where no issue under Windows 7.


    There are no DX12 games yet for review, so why would you assume that you should see better performance in games made for windows 7 DX11, in win10 DX12? Especially in "tri-Fire". DX12 has significant advantages over DX11 so you should wait till these games actually come out before making assumptions on performance, or the validity of DX12's ability to increase performance.
    My games, FO4, GTAV and others run better in Win10 and i have had zero problems. I think your issue is more Driver related, which is on AMD's side, not MS's operating system.
    I'm on the Red team by the way.
    8
  • ohim
    Quote:
    Quote:
    For someone who runs Crossfire R9 390x (three cards) DX12 makes no difference in term of performance. For all BS Windows 10 brings not worth *downgrading to considering that lot of games under Windows 10 are simply broken or run like garbage where no issue under Windows 7.


    There are no DX12 games yet for review, so why would you assume that you should see better performance in games made for windows 7 DX11, in win10 DX12? Especially in "tri-Fire". DX12 has significant advantages over DX11 so you should wait till these games actually come out before making assumptions on performance, or the validity of DX12's ability to increase performance.
    My games, FO4, GTAV and others run better in Win10 and i have had zero problems. I think your issue is more Driver related, which is on AMD's side, not MS's operating system.
    I'm on the Red team by the way.
    Win 10 and AMD card and no issues here.
    7
  • chimera201
    Such research. Much detail. Nice benchmark analysis TH.
    10
  • FormatC
    I've migrated all 8 machines here to Windows 10 (Windows 7 und 8.1 images are saved for my museum) - no problems since months. With a good hosts file all my PCs can't call home to MS, a must-have in each case. Windows 10 runs stable, mostly faster on my six- and eight-cores and I also have no issues with porper programmed old applications from the 90's (32 bit). The only issue are changed rights in a lot of registry keys, but this can be changed manually (with trusted installer rights :P )
    3
  • Math Geek
    what i found most interesting is the cpu bottleneck on even the high end cpu used in the test system. i REALLY want to see this same test done on a 4 core i5 cpu to see what the loss of 4 cores does to the numbers. this is likely what many users have at home and not the super set-up used here.

    other thought i had was what this would look like on an fx-8*** system. we know that if you tax the fx cpu on all cores it will out perform an i5 due purely to the extra resources it has. so i wonder how bad the bottleneck will be on an i5 and an fx cpu and speculate that the fx chip might actually get a second life in this scenario. might pull even or even pull ahead of an i5 in this benchmark.

    in any case, i'd love to see the numbers run again with these cpu's just to see what happens.
    5
  • Kahless01
    so you have to turn async on through the ini? I downloaded this Saturday and running their in game benckmark says im gpu bound by a lot. about 70% on average. im running a 380x and i7-920. plays like total annihilation. I like it. if I can get a little better performance turning async on ill take it. I do get slowdown when I use the bombers and dreadnought launching missles over the horizon.
    2
  • FormatC
    You must start it in DX12 mode. Async is in this case always on (per default). This INI entry hasn't any effect in DX11 mode. I need two high-end cards to run into CPU bound (5960X @4.8 GHz) with my private rig. I get approx. 10% of extra performance without SMT enabled (only 8 threads).
    4
  • jimmysmitty
    Anonymous said:
    Quote:
    An AMD sponsored title
    Really? Sponsoring and knowledge sharing are two pairs of shows. Nvidia was invited too. ;)

    Quote:
    Async Compute will quickly start to take the same road taken by Mantle

    Sure? The design of the most current titles and engines was started long time before Microsoft started with DirectX 12. You can find the DirectX 12 render path in first steps now in a lot of common engines and I'm sure that PhysX will faster die than DirectX12. Mantle was the key feature to wake up MS, not more. And: it's async compute AND shading :)


    He is not talking about DX12 dying but rather that ASYNC Compute will and unless games heavily adopt it it might be a one off feature.

    That said, I would not be surprised if nVidia adopts it in Pascal as it is a feature that might also benefit workstation rendering.

    Hard to say what will happen in the long run but I think Polaris and Pascal will matter more for DX12 than Fury or the 900 series.
    0
  • 17seconds
    That was exactly my point, jimmy.

    This game uses the exact same engine developed by Oxide for the Star Swarm demo. Star Swarm was developed as the first benchmark to showcase AMDs Mantle technology.

    Now the same development team with the same game engine has released the first demo to take advantage of AMDs Asynchronous Compute capabilities. As with the Star Swarm demo, this is a best case scenario for AMD. The parallels between the hype generated by Oxide for Async Compute and Mantle are hard to ignore. In the end, the results will be the same if only a few games adopt the technology.
    -5
  • NightAntilli
    A few statements...

    The idea that the Fury line is more power hungry than the 980 series has been disproved in this TH benchmark.

    All mid range and higher AMD cards have been running between ~50% - ~75% of their true capabilities due to front end limitations under DX11. DX12 can bypass these, showing the true compute power of the cards.

    nVidia CANNOT compensate for async compute with a software update. Their hardware is simply incapable. Their necessary pre-emption holds them back. Do not get your hopes up.

    Pascal will not likely solve this issue. We won't be seeing a properly working async implementation from nVidia until Volta (Pascal's successor).

    Async compute is getting implemented in consoles regularly. DX12 will be using this, because it enables developers to do more with the same hardware. Multiple titles will be released this year. The next one is Hitman. Wait for the DX12 benchmarks in March... Don't be surprised if the R9 390 (non-x) reaches 980 Ti levels.

    The performance gain from AMD here is representative of all DX12 titles that use async compute. In many cases, it will even be better.

    Async compute cannot work under DX11, which is why it appeared that nVidia cards were the better choice. For long term, GCN is and always has been the better choice. Sad to say, but Maxwell 2 cards from 2015 are already more outdated than GCN 1.1 cards from 2013.

    Final statement is, do not believe me. You never do anyway, so no reason to change now. Instead, I ask you to wait and see, and you will find out for yourself soon enough.
    6
  • jimmysmitty
    Anonymous said:
    A few statements...

    The idea that the Fury line is more power hungry than the 980 series has been disproved in this TH benchmark.

    All mid range and higher AMD cards have been running between ~50% - ~75% of their true capabilities due to front end limitations under DX11. DX12 can bypass these, showing the true compute power of the cards.

    nVidia CANNOT compensate for async compute with a software update. Their hardware is simply incapable. Their necessary pre-emption holds them back. Do not get your hopes up.

    Pascal will not likely solve this issue. We won't be seeing a properly working async implementation from nVidia until Volta (Pascal's successor).

    Async compute is getting implemented in consoles regularly. DX12 will be using this, because it enables developers to do more with the same hardware. Multiple titles will be released this year. The next one is Hitman. Wait for the DX12 benchmarks in March... Don't be surprised if the R9 390 (non-x) reaches 980 Ti levels.

    The performance gain from AMD here is representative of all DX12 titles that use async compute. In many cases, it will even be better.

    Async compute cannot work under DX11, which is why it appeared that nVidia cards were the better choice. For long term, GCN is and always has been the better choice. Sad to say, but Maxwell 2 cards from 2015 are already more outdated than GCN 1.1 cards from 2013.

    Final statement is, do not believe me. You never do anyway, so no reason to change now. Instead, I ask you to wait and see, and you will find out for yourself soon enough.


    It is hard to trust any one person on a forum vs what we have seen.

    One thing you forgot to remember, the PS4 is not using DX12. It uses its own version of OpenGL since it uses a Linux based kernal. The XB1 will use DX12 but from everything I have already read it wont benefit nearly as much as the PC will since the way that the XB1s DX is written currently it already takes better advantage of that hardware that is in there. So even the ASYNC might not benefit the XB1.

    Another factor is that DX12 is more than just ASYNC. ASYNC is just one aspect, much like Tesselation was just one aspect of DX11. It is not the end all be all. For all we know, this specific benchmark was optimized solely for ASYNC work and not for other aspects which could make a massive difference.

    Since the consoles still have their differences on the API level that could influence game developers and for PC game developers will have to look at the market, which right now is heavily nVidia based.

    If AMD really wants ASYNC to take off then they need to do what nVidia is doing and not what they normally do. They normally partner with one company to show case a tech instead of trying to partner with multiple companies. Mantle was mostly shown off in BF4. The only other major title I remember it in was Thief and it only really benefited systems with super high end GPUs and low end CPUs.

    I still say this is in no way a 100% picture. We need games that are developed for the purpose of entertainment, not for the purpose of benchmarking.
    0