Announcement

Collapse
No announcement yet.

NVIDIA GeForce GTX 1080 Ti: Windows 10 Creators Update vs. Ubuntu Linux Gaming

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • NVIDIA GeForce GTX 1080 Ti: Windows 10 Creators Update vs. Ubuntu Linux Gaming

    Phoronix: NVIDIA GeForce GTX 1080 Ti: Windows 10 Creators Update vs. Ubuntu Linux Gaming

    Earlier this week I posted some fresh AMD Radeon Windows vs. Linux gaming benchmarks using the newly-released Windows 10 Creator Update and Ubuntu 17.04. For your viewing pleasure today are some fresh NVIDIA Windows vs. Linux benchmarks using a high-end GeForce GTX 1080 Ti graphics card.

    Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite

  • #2
    Well, thats what I call a nice reference point. When looking at the AMD results, one could only guess if it is the driver or a bad port. With these results, we pretty much see which impact the porting has and to which performance the OS-drivers can aim at, if they want to reach parity.

    The sanity check towards "unported" games shows, that Linux is not at fault for the reduced performance and with that I rather like the results.
    Last edited by Shevchen; 28 April 2017, 07:52 AM. Reason: typo

    Comment


    • #3
      I wouldn't call those ports "bad" per se. They require a lot of work, and as long as they are playable, I am glad to have them.

      Now, they might not deliver as well as a game designed with cross-platform in mind, performance-wise, and that't fully expected. It's a port, after all, not a game designed from the ground up for the platform.
      The fact that those games are already a few months/years old also usually helps with the performance.
      But of course, I agree that it looks bad on the benchmarks, and that I wouldn't mind a bit of extra performance.

      But let's make up for it by speeding up our open source drivers! I agree with Shevchen, AMD's results look even more impressive when compared to these.

      Comment


      • #4
        Would have _really_ enjoyed seeing how different Dota 2 results would've been. Dota 2 on linux has been running very well, and it's very much a 1st class citizen.

        Comment


        • #5
          Originally posted by Shevchen View Post
          Well, thats what I call a nice reference point. When looking at the AMD results, onc could only guess if it was the driver or a bad port. With these results, we pretty much see which impact the porting has and to which performance the OS-drivers can aim at, if they want to reach parity.

          The sanity check towards "unported" games shows, that Linux is not at fault for the reduced performance and with that I rather like the results.
          It's what keeps me in the green camp: whatever performance I get on Windows, I get on Linux, too.
          The other reason is I can upgrade Ubuntu without fear of an unsupported kernel/X. I know AMD has fixed that, but I've been enjoying this freedom for about decade now.

          Edit: that's for my home system. For work (sw dev) I can get by with any IGP.
          Last edited by bug77; 28 April 2017, 06:37 AM.

          Comment


          • #6
            Originally posted by M@yeulC View Post
            I wouldn't call those ports "bad" per se. They require a lot of work, and as long as they are playable, I am glad to have them.
            Sure, the performance difference is not that big that you would disregard the game. As long as you get 60 FPS you are fine. The thing is, that the reference point here is the really interesting part.

            Originally posted by M@yeulC View Post
            Now, they might not deliver as well as a game designed with cross-platform in mind, performance-wise, and that't fully expected. It's a port, after all, not a game designed from the ground up for the platform.
            The fact that those games are already a few months/years old also usually helps with the performance.
            But of course, I agree that it looks bad on the benchmarks, and that I wouldn't mind a bit of extra performance.

            But let's make up for it by speeding up our open source drivers! I agree with Shevchen, AMD's results look even more impressive when compared to these.
            And this is the main beef here - If I subtract the percentages from the port and apply them to the AMD drivers, suddenly they look quite good. They are still lacking behind a bit - but this is becoming a minor issue and the greater impact right now is the port itself, not the driver being bad.

            Originally posted by bug77 View Post

            It's what keeps me in the green camp: whatever performance I get on Windows, I get on Linux, too.
            The other reason is I can upgrade Ubuntu without fear of an unsupported kernel/X. I know AMD has fixed that, but I've been enjoying this freedom for about decade now.

            Edit: that's for my home system. For work (sw dev) I can get by with any IGP.
            Well, for me its the other way around. The more I look at the relative performance, the more I like the OS-AMD drivers. I'm pretty confident, that within the next half year or so, there will be next to no big issue left for the OS-drivers and that it can focus purely on optimization, refinement and performance increases. Esp. when all Open GL 4.5 functions are set and done, we can shift the focus heavily towards RADV and maybe even get tools done like Wattman for fine tuning performance, instead of hacking them into the system and hope for the best.

            But I can see your point of view too. Nvidia gives your reliable performance and you don't have to be afraid of severe regression. Thing is, the last couple of months, Nvidia did not earn roses for their (Windows) drivers, while AMD did... a little. So, for a future purchase - those results give me a good feeling about buying Vega and for the green team its also a save thing.

            Good performance on both sides has only advantages for us end-users. But as I want to rock with Gallium 9, my decision is already made. No hard feelings.
            Last edited by Shevchen; 28 April 2017, 07:54 AM.

            Comment


            • #7
              Originally posted by Shevchen View Post
              Sure, the performance difference is not that big that you would disregard the game. As long as you get 60 FPS you are fine. The thing is, that the reference point here is the really interesting part.


              And this is the main beef here - If I subtract the percentages from the port and apply them to the AMD drivers, suddenly they look quite good. They are still lacking behind a bit - but this is becoming a minor issue and the greater impact right now is the port itself, not the driver being bad.



              Well, for me its the other way around. The more I look at the relative performance, the more I like the OS-AMD drivers. I'm pretty confident, that within the next half year or so, there will be next to no big issue left for the OS-drivers and that it can focus purely on optimization, refinement and performance increases. Esp. when all Open GL 4.5 functions are set and done, we can shift the focus heavily towards RADV and maybe even get tools done like Wattman for fine tuning performance, instead of hacking them into the system and hope for the best.

              But I can see your point of view too. Nvidia gives your reliable performance and you don't have to be afraid of severe regression. Thing is, the last couple of months, Nvidia did not earn roses for their (Windows) drivers, while AMD did... a little. So, for a future purchase - those results give me a good feeling about buying Vega and for the green team its also a save thing.

              Good performance on both sides has only advantages for us end-users. But as I want to rock with Gallium 9, my decision is already made. No hard feelings.
              WHEN AMD will get there, I'll have no problem switching (I've used ATI and AMD before). But the point I was trying to make, is Nvidia got there 10 years ago.
              Also, what hard feelings? I've always said users should use whatever they want. Forums are for discussing pros and cons, not for making choices for others

              Comment


              • #8
                Originally posted by FishPls View Post
                Would have _really_ enjoyed seeing how different Dota 2 results would've been. Dota 2 on linux has been running very well, and it's very much a 1st class citizen.
                I agree with that, will be a nice dota 2, it's a game been active than anyone in all platforms.

                Comment


                • #9
                  Originally posted by bug77 View Post

                  WHEN AMD will get there, I'll have no problem switching (I've used ATI and AMD before). But the point I was trying to make, is Nvidia got there 10 years ago.
                  Also, what hard feelings? I've always said users should use whatever they want. Forums are for discussing pros and cons, not for making choices for others
                  The "no hard feeling" part is meant to show my neutral opinion. I don't really care about a brand, I want stuff that works. I have a couple of games that must run under Wine (DX9) and for a halfway decent performance, this needs gallium9. AMD supports it, Nvidia doesn't. IF Nvidia would support it, the game would change.

                  Other than that, I want to see if the tech AMD announced for Vega is really working out - so in any case I'll wait until I get a couple of benchmarks in and will not blindly buy it.

                  Comment


                  • #10
                    Originally posted by Shevchen View Post
                    As long as you get 60 FPS you are fine.
                    Well, it's cool to have more when you have a 144Hz monitor. 🙂

                    Comment

                    Working...
                    X