Announcement

Collapse
No announcement yet.

Mesa 17.1-dev vs. AMDGPU-PRO 16.60 vs. NVIDIA 378 Linux Gaming Tests

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

  • #31
    Originally posted by dungeon View Post

    Or don't forget you might want some CPU with iGPU or these with eDRAM Well i dunno, these much bellow $100 GPUs are always combined with much cheaper than $200 CPUs as if you go up $300 for both you might wanna forget about very cheap dGPU option

    Sounds irregular that Michael's Xeon or even that AMD's i7 6700K to be combined with RX 460 That might be only for marketing provocations reasons as that dGPU is faster than *any* iGPUs ... as whatever iGPU you have does not matter if CPU cost $200 or $2000, that RX 460 is still better so applies anywhere
    I know, I'm just (basically) saying that from my perspective it doesn't matter which CPU he uses, as long as it is fast enough to show which frame rate the tested GPU can achieve with that driver. This assumes the frame rate won't depend much on the specific CPU, as long as the CPU can keep the data coming. To figure out which CPU is the cheapest that can provide the required data flow (needed for a specific game/card/driver/resolution/frame-rate combination), that is a second step.

    Comment


    • #32
      Originally posted by bridgman View Post

      I'm not seeing it (at least relative to RX 480) - the 460 has <40% of shader power and ~50% of everything else, and it's running around 45% the speed of the 480. There is one exception where the difference is more, guessing that is memory size related with fewer heuristics in the open source stack for not-quite-enough-VRAM scenarios.

      I agree that 460 should be faster than 260 in all cases though.

      And CAN YOU TURN OFF THOSE DAMN SMILIES ?
      So, I hear you like smilies....



      Sorry. Had to.

      Comment


      • #33
        Originally posted by debianxfce View Post

        Yep and waste 40 euros, Gigabyte RX460 2Gb vs gtx1050. RX460 has more transistors and is faster with DX12 and Rise of the Tombraider. Tombraider benchmark is all what matters and with wine-staging and TR2013 I have max 100 fps avg 75 fps at 1920x1200 normal settings. Also phoronix benchmarks are out of real world with 640 usd cpu.
        No it doesn't. RX460 3bn, 1050 3.3bn

        http://videocardz.net/his-radeon-rx-...im-icooler-oc/



        Don't let your fanboyism get in the way of facts. I don't know where you are buying your 1050s from but you are getting ripped off, the difference is not 40 euros.

        Comment


        • #34
          Thanks for the benchmarks, really exciting numbers! First up, it is clear that radeonsi is now more than competitive with the amdpro driver. Vulcan of course has a way to go, and Dirt is the lone exception on the OpenGL side. Hopefully we will see both fixed this year. Focusing on the Fury card, comparing it to 980 and skipping the Vulcan results, I get the following:
          Bioshock 50% (strong win for nvidia, but seems limited by something with radeonsi)
          CS:GO 108% (a win for radeonsi!)
          Deus Ex 84%
          Dirt 29% (Ouch, desperately needs some optimization in Radeonsi)
          Dota 2 HD 78%
          Dota 2 4K 89%
          Metro 75%
          TF2 93%
          ET:LE 87%
          OpenArena 64%
          Tesseract 98%
          Unigine Heaven 87%
          Xonotic 101% (win number two for radeonsi)

          To sum up, Dirt needs some work. OpenArena is not really relevant for high end cards. Radeonsi is now within striking distance, even winning two over the proprietary nvidia driver. Keep it up guys, you are winning!

          Comment


          • #35
            Originally posted by bridgman View Post

            (answering Dungeon)

            And CAN YOU TURN OFF THOSE DAMN SMILIES ?
            Michael : you should charge the smileys 1cent and make Phoronix a cash machine thanks to Dungeon

            Comment


            • #36
              Originally posted by Del_ View Post
              Thanks for the benchmarks, really exciting numbers! First up, it is clear that radeonsi is now more than competitive with the amdpro driver. Vulcan of course has a way to go, and Dirt is the lone exception on the OpenGL side. Hopefully we will see both fixed this year. Focusing on the Fury card, comparing it to 980 and skipping the Vulcan results, I get the following:
              Bioshock 50% (strong win for nvidia, but seems limited by something with radeonsi)
              CS:GO 108% (a win for radeonsi!)
              Deus Ex 84%
              Dirt 29% (Ouch, desperately needs some optimization in Radeonsi)
              Dota 2 HD 78%
              Dota 2 4K 89%
              Metro 75%
              TF2 93%
              ET:LE 87%
              OpenArena 64%
              Tesseract 98%
              Unigine Heaven 87%
              Xonotic 101% (win number two for radeonsi)

              To sum up, Dirt needs some work. OpenArena is not really relevant for high end cards. Radeonsi is now within striking distance, even winning two over the proprietary nvidia driver. Keep it up guys, you are winning!
              Thanks for this info, saved me time! Maybe this percentage could be considered as progress towards "mission complete" status. This based on the assumption that nvidia blob is representative of highest realistically achievable performance.

              Comment


              • #37
                Originally posted by Xicronic View Post
                Is Shadow of Mordor going to be added to the testing suite? That Feral representative from the other thread said there was a CLI option to launch it.
                Nope. Heard back finally....
                Sorry for the delay - we were checking with our technical team to make sure the information is correct.

                We can confirm Mordor doesn't have any publicly-listed automated benchmark options on any platform right now.

                We have been looking into if something could be officially tested and documented for Shadow Of Mordor in the future, however for now this game does not have an official command line benchmark option.

                We're sorry for confusion in your forum about this, we do benchmark a number of titles internally and our staff member was mistook Mordor for a different title.
                Michael Larabel
                https://www.michaellarabel.com/

                Comment


                • #38
                  Originally posted by oooverclocker View Post

                  Tile Based Rendering is a software implementation dividing one frame into pieces being rendered simultaneously to achieve higher usage. AMD uses TBR on Windows as well, as far as I know. And when the game engine supports it it also works with OpenGL. Sadly TBR is not infinitely scalable.
                  Go back and check out what i actually linked to. Tile based rasterization. It's a hardware feature.

                  See the video here for details: http://www.realworldtech.com/tile-ba...n-nvidia-gpus/

                  Apparently Vega supposedly introduces the same thing into AMD hardware, but I'm not sure if that's really confirmed or still just a rumor at this point. Their current GPUs definitely don't have it though.
                  Last edited by smitty3268; 01 February 2017, 02:44 AM.

                  Comment


                  • #39
                    Well I get a different picture on my RX 480 so the whole process is definitely different on GCN GPUs than with the old 6670 from the video but it still doesn't differ as much as Nvidias 900+ GPUs.

                    Comment


                    • #40
                      Originally posted by debianxfce View Post
                      Gtx 1050 was not available when rx460 was in sale but gtx 950 is.

                      Nvidia cards are more expensive and you need to kill X when installing the nvidia driver. When building a pc, cost and easy to use are the most important factors for all major builders.

                      Keep back-pedalling dude



                      Healey's first law of holes:

                      When you find yourself in a hole, stop digging.

                      Gigabyte RX460 2Gb vs gtx1050. RX460 has more transistors

                      Comment

                      Working...
                      X