Announcement

Collapse
No announcement yet.

The AMD Radeon R9 Fury Is Currently A Disaster On Linux

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

  • #71
    The card performance lies within expectation: Better than R9 290 on OpenGL but still fall far behind NVIDIA card at even lower price. Top performance on OpenCL.
    It's disaster from Michael point of view: nevertheless he paid $600 to this.

    Given current situation it's not a good idea to purchase this card for OpenGL/3D application/games with closed-source driver on Linux. NVIDIA is a better choice.
    But hey, that probably has never been the reason for people to buy AMD card for Linux, right?
    Either they bought them for good/decent open source support for indie gaming, or supreme OpenCL with closed driver.

    I will still keep R9 Nano on my wish list for Christmas, if the open-source driver support becomes good by then.
    Very likely a MSI one as they seems to be kind enough to send sample to Michael.

    Comment


    • #72
      Originally posted by Michael View Post

      Do you expect any OpenArena developers to be at the event? It's been more than three years since their last release, I'm sure they have a big one coming up soon.....

      Your OpenArena comments still make no sense even after I posted those few R9 Fury ones at sub-4K... I don't think I'm alone either given the comments by others.
      If you think that it does not made sense why don't you benchmarking TF2 then, it has profile... any game that has profile

      I mean really this is biased article Michael... no idea about OA devs, and i really wants to be a bit informed about W10 on its launch day, mostly reading those today.

      Comment


      • #73
        Originally posted by dungeon View Post

        If you think that it does not made sense why don't you benchmarking TF2 then, it has profile... any game that has profile

        I mean really this is biased article Michael... no idea about OA devs, and i really wants to be a bit informed about W10 on its launch day, mostly reading those today.
        I'm so irked about that. This afternoon I got slammed on windows update problems. I got a few on the bench now and it's looking like the windows 10 upgrade is having trouble on systems that are bogged down with useless background processes. I highly suggest that background processes get stopped before the upgrade attempt. Do a cleanup on the existing windows first.

        It kinda sucks that MS pushed it out via WU. There are too many screwed up systems out there to assume that it'll go smoothly.

        EDIT: The only reason I can think of for MS to push this upgrade via WU is that it incorporates something that MS wants on as many machines as possible. I don't know what that thing might be, but I'll bet real money it's malicious.
        Last edited by duby229; 30 July 2015, 02:00 AM.

        Comment


        • #74
          Maybe MS is trying to start a new business after failure in mobile markets. Maybe this new business is based on data, where the biggest competitors are Facebook and Google

          Comment


          • #75
            Originally posted by dungeon View Post

            It is not Catalyst problem for openarena, did you readed anything from links that you quoted?

            http://openbenchmarking.org/prospect...edf1fa68b1445d

            R9 290 had 263fps and GTX 980 has 295fps... that is around 10% difference (that is not any big bottleneck for any driver as you see) , and Fury's Michael results are slower then both now even on a bit lower resolution.
            Dungeon will you one day understand that FPS differences above 100 means NOTHING about power.
            Drivers and GLs are optimized for 30 -> 60 fps.

            Please run yourself OpenArena at 640x480 or Doom 1 at 320x240 to benchmark your CPU/GPU and get stupid results.

            Oh, before I forget, I agree with Duby, we don't care about memory bandwidth. If it does not give some performance boost then it is just a marketing number.

            Comment


            • #76
              Originally posted by dungeon View Post

              It is not simple as that - results are a bit more biased then a expect, my Micheal... nearly all games you tested have nvidia variabile in game scripts and not a single game (other then tesseract i think) has fglrx profile. Unigine clearly shows where GPU is, those has profiles. Disaster? Well, no

              Please test TF2 and openarena
              Basically, the only ways to put AMD into a decent light are a title from 2007 and a title that's work in progress and hasn't seen a release since 2012. Yeah, that seems like a great way to put a $600 video card to use. And OpenCL, but that's a niche segment right now.
              Last edited by bug77; 30 July 2015, 06:25 AM.

              Comment


              • #77
                Originally posted by bug77 View Post

                Basically, the only way to put AMD into a decent light are a title from 2007 and a title that's work in progress and hasn't seen a release since 2012. Yeah, that seems like a great way to put a $600 video card to use. And OpenCL, but that's a niche segment right now.
                Please Michael benchmark PONG with your brand new Fury for Dungeon as is it memory intensive and as an AMD profile.
                I am sure it will beat a Titan X by 2% (1515471123 fps vs 1493509236) for 400$ less so that everyone can see how AMD is superior.
                In your face NVidia!!!

                Comment


                • #78
                  Wow...Valve really has put pressure on the GPU manufacturers eh

                  Comment


                  • #79
                    Hi Michael & yall,

                    great article mate, keep it up. Just ignore the white noise and filter for signal.

                    There is no convincing "the flat earth society," about scientifically repeatable experiments.... So much then, for astro-turfing trolls.

                    As Anton points out here, "
                    http://www.kitguru.net/components/graphic-cards/anton-shilov/amd-vows-to-keep-rd-investments-at-appropriate-levels/" nVida spend about 30% on R&D, whereas ATI/amd spend less that 25% of income on R&D. Also, nVida's income is higher. Guess whose product is better?

                    Now, I am no rocket scientist, but if ATI management spent more on product & less on BS, do you think there product would be better???? Ya know, more peeps might even buy it if they did so.... As it stands & has done for about 14yrs, the accepted Linux Wisdom, if you want performance, use nVida. ATI management are responsible for the ongoing debacle.

                    Anyone wanna shout them a Coffee or maybe some Bavarian Ale? ;-)

                    GreekGeek :-)

                    Comment


                    • #80
                      New, or any, AMD GPU's being a disaster on Linux is not news anymore - its just a matter of fact. Please save the bandwidth and write a article when a AMD GPU performs well on Linux - that is something that's newsworthy.
                      Last edited by teeedubb; 30 July 2015, 06:42 AM.

                      Comment

                      Working...
                      X