Announcement

Collapse
No announcement yet.

Running The NVIDIA GeForce GTX 680 On An Open-Source Driver

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

  • Running The NVIDIA GeForce GTX 680 On An Open-Source Driver

    Phoronix: Running The NVIDIA GeForce GTX 680 On An Open-Source Driver

    Thanks to clean-room reverse-engineering, it is already possible to run the NVIDIA GeForce GTX 680 "Kepler" graphics card on a fully open-source graphics driver complete with OpenGL acceleration. Here are the first benchmarks of this work-in-progress, community-created open-source GeForce 600 series graphics driver.

    http://www.phoronix.com/vr.php?view=17296

  • #2
    I glanced through the article twice but did not see clock speed listed. Can you please add the clock speed(s) to the comparison chart so that we may extrapolate what the performance might be when re-clocking is supported?

    F

    Comment


    • #3
      Why as unity 2d used for nouveau?

      Comment


      • #4
        for me the open-source driver for nvidia hardware is a impressive miracle.

        how is this possible ? we get faster open-source support without any support from nvidia than we get open-source support from the newest hd7970 amd card with official open-source support.

        maybe the VLIW architecture is the worst case for an open driver?

        edit: "Two of the Nouveau developers managed to get their hands on two GeForce GTX 680 graphics cards before launch. As a result, and since the NVIDIA driver changes from the GeForce 400/500 "Fermi" to GeForce 600 "Kepler" series were not too invasive, they managed same-day open-source driver support."

        Damn these guys are good.. amd's show is really sad compared to this.

        these hacker nerds get same-day-opensource-driver-support and the official supported amd driver for the hd7970 don't run month after selling the cards to every street bums... sad... really sad
        Last edited by Qaridarium; 04-26-2012, 03:27 PM.

        Comment


        • #5
          "clean-room reverse-engineering"

          what? dude, just call it reverse engineering.

          Comment


          • #6
            that performance is painful

            Comment


            • #7
              Originally posted by Qaridarium View Post
              for me the open-source driver for nvidia hardware is a impressive miracle.

              how is this possible ? we get faster open-source support without any support from nvidia than we get open-source support from the newest hd7970 amd card with official open-source support.

              maybe the VLIW architecture is the worst case for an open driver?

              edit: "Two of the Nouveau developers managed to get their hands on two GeForce GTX 680 graphics cards before launch. As a result, and since the NVIDIA driver changes from the GeForce 400/500 "Fermi" to GeForce 600 "Kepler" series were not too invasive, they managed same-day open-source driver support."

              Damn these guys are good.. amd's show is really sad compared to this.

              these hacker nerds get same-day-opensource-driver-support and the official supported amd driver for the hd7970 don't run month after selling the cards to every street bums... sad... really sad
              SI (radeon hd 7xxx) is a completely new architecture so very little of code from older asics can be re-used (SI is no longer VLIW like previous asics). Kepler is just an evolutionary update to fermi so there are a lot fewer changes to deal with in the driver. Note that Trinity is an updated version of an existing architecture. Trinity hardware has not yet been released, but there is already full support for it in the open source driver.

              Comment


              • #8
                Originally posted by agd5f View Post
                Trinity hardware has not yet been released, but there is already full support for it in the open source driver.
                A nice full support slow as hell


                @Michael
                PLEASE TEST TRINITY WITH POWER PROFILE HIGH
                ## VGA ##
                AMD: X1950XTX, HD3870, HD5870
                Intel: GMA45, HD3000 (Core i5 2500K)

                Comment


                • #9
                  Originally posted by agd5f View Post
                  SI (radeon hd 7xxx) is a completely new architecture so very little of code from older asics can be re-used (SI is no longer VLIW like previous asics). Kepler is just an evolutionary update to fermi so there are a lot fewer changes to deal with in the driver. Note that Trinity is an updated version of an existing architecture. Trinity hardware has not yet been released, but there is already full support for it in the open source driver.
                  i know all this but: the catalyst also deal with the same problem and the catalyst driver is also ready on release date!
                  the catalyst driver prove one think about the hd7xxx series all your arguments are just excuses !
                  the catalyst driver prove it! you can build a driver on release date!

                  the open-source driver is only not ready "yet" because amd think the open-source driver is a play-toy for nerds only useful for latency support excuses!

                  they just save "money" on the open-source side to make sure Microsoft is happy means they can earn a lot of money for the next Xbox hardware chips.

                  I'm sure for the Xbox there are no excuses there are drivers on release date!

                  and about "Trinity" Trinity is "Low-end" your words translated to the reality means AMD think only low-end customers use the opensource driver.

                  Comment


                  • #10
                    Evolutionary

                    Sure it is very similar to Fermi, it's still a "D3D11 card" (.1 doesn't change much), there's no reason to change anything major about the interface to the 3D engine. The shader ISA hasn't changed a lot either, but instruction scheduling has become more important for performance.

                    About clocks, we can't decode the current speeds yet, but it's probably the 1st performance level the blob uses - decent GPU speed but very low memory clock (about 1/10th, so that should be the first bottleneck).

                    Btw., that 2nd GTX680 I actually just bought on release day.

                    Comment

                    Working...
                    X