Announcement

Collapse
No announcement yet.

Intel HD 4000 Ivy Bridge Graphics On Linux

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

  • #71
    Originally posted by Kano View Post
    Just take the PSU from another pc or is it the one and only desktop in the house...
    Why can't you just get it through your damned head that he's looking for an excuse to upgrade?!

    Comment


    • #72
      Originally posted by Tgui View Post
      Why can't you just get it through your damned head that he's looking for an excuse to upgrade?!

      Because I have tools and so can you!

      Comment


      • #73
        Originally posted by baffledmollusc View Post
        The thing that impresses me here is how good the open-source Intel drivers are, in performance, at least.
        Thanks! Many people worked really hard on it for many months, both within Intel and in the open-source community, to get to this point. That is the power of open-source!

        And also, speaking about performance, just wait a few months for Kernel 3.4 and 3.5, each should come with a very nicely looking set of performance-related patches for both Sandy and Ivy Bridge.

        Comment


        • #74
          Impressive. CPU and GPU performance is great, Linux support awesome, and this should make really good cpu for laptops. Waiting for OpenCL and video acceleration support (VDPAU preferably ;D ). Intel Linux team is doing great job.

          Best x86 processor ever, need to buy it quickly . Hope AMD will respond properly.

          Comment


          • #75
            @eugeni_dodonov

            When can we expect stability fixes for xserver 1.11.x and newer hat kde 4.x does not crash the xserver or get artefacts when you use disable composite for fullscreen opengl apps?

            Comment


            • #76
              Originally posted by Kano View Post
              @eugeni_dodonov

              When can we expect stability fixes for xserver 1.11.x and newer hat kde 4.x does not crash the xserver or get artefacts when you use disable composite for fullscreen opengl apps?
              Is there any bug about this by a chance? I don't recall any similar issues that I've seen on bugzilla lately..

              Also, what GPU card is it? Sandy Bridge? And what is the xf86-video-intel and mesa version where it happens?

              Comment


              • #77
                It happens with snb and ivb on pure debian wheezy or sid.

                Comment


                • #78
                  "Holy crap" is my initial impression of the HD 4000 graphics on my 3770K. It's release week (it was only offered up for sale at Amazon over the weekend I believe); I have just finished installing my new system; I grabbed the latest updates from Fedora 17 Beta, and WOW. It FLIES. Here are the highlights:
                  • 30 fps in Second Life (64-bit native client, Armin Weatherwax's Teapot custom viewer)
                  • Perfectly fluid Kwin compositing
                  • 2D performance visually appears to beat Catalyst by a mile, or at least the responsiveness to input events -- you move the mouse and it just goes. Scrolling in web browsers is amazing. It's like using an iPhone, except open source and a desktop. My main complaint about Catalyst is that it has terrible tearing artifacts with Tear-Free disabled, and horrible responsiveness with Tear-Free enabled. You get both with Intel. I don't know how. r600g gives you low responsiveness and no artifacts also, but I can't use r600g with my SI card.
                  • Haven't found any rendering artifacts, instabilities or indications of poor performance yet. The only things I miss are OpenGL 3.0 (which only appears to be available for Sandy Bridge right now, at least on Fedora 17 Beta) and of course the non-existent hardware-accelerated OpenCL.


                  I am really excited that Intel was able to support Ivy Bridge so far in advance, and the support is so robust on release week. Thanks to all the Intel folks and community members who contributed to this; it makes a real difference for those of us who need working and stable open source graphics drivers

                  AMD, TAKE NOTES: your competitors are doing much better than you with the stability and performance (or in some cases, mere existence) of open source driver support on release day of their products. The Intel drivers for Ivy Bridge are, as far as I'm concerned, a "grand slam home run". The NI-derived APU open drivers I would say are a "single" base-hit. The SI discrete / Trinity APU open drivers I would say are a "strike-out" or "ground into double play". If you are looking for an open source driver release schedule that is considered acceptable to "picky" customers (I would consider myself picky), all you have to do is emulate what Intel just did with Ivy Bridge. Anything more than that is just icing on the cake.

                  Comment


                  • #79
                    Originally posted by baryluk View Post
                    Impressive. CPU and GPU performance is great, Linux support awesome, and this should make really good cpu for laptops. Waiting for OpenCL and video acceleration support (VDPAU preferably ;D ). Intel Linux team is doing great job.

                    Best x86 processor ever, need to buy it quickly . Hope AMD will respond properly.
                    IVB should have vaapi acceleration, just like snb, except better.
                    I doubt we'll see openCL on linux for awhile since they can't leverage the gallium work, and would thus have to produce it from scratch on their own (which they could easily do, but I doubt it is a priority).

                    Comment


                    • #80
                      Originally posted by allquixotic View Post
                      AMD, TAKE NOTES: your competitors are doing much better than you with the stability and performance (or in some cases, mere existence) of open source driver support on release day of their products. The Intel drivers for Ivy Bridge are, as far as I'm concerned, a "grand slam home run". The NI-derived APU open drivers I would say are a "single" base-hit. The SI discrete / Trinity APU open drivers I would say are a "strike-out" or "ground into double play". If you are looking for an open source driver release schedule that is considered acceptable to "picky" customers (I would consider myself picky), all you have to do is emulate what Intel just did with Ivy Bridge. Anything more than that is just icing on the cake.
                      I wish they (AMD) read this.

                      Comment

                      Working...
                      X