Announcement

Collapse
No announcement yet.

Open-Source Graphics Driver Updates For Linux 5.11 Have Lots Of Intel + AMD Presents

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

  • #11
    Originally posted by zexelon View Post
    I must say I am a happy Nvidia proprietary driver user... I dont have to hack, cut, compile or beg Linux to run with it, it just works. As an added bonus I also get hardware that tends to kick AMD in the teeth. Though this latest gen from AMD is much more impressive and I explored converting, but unfortunately they still have nothing compared to CUDA and thats where I spent the vast majority of my time these days.
    If the goal is just to get an accelerated desktop GUI on Wayland, Nouveau does well enough as long as the cards are at least two generations older and Plasma Wayland is not used.

    And as far as I am concerned, Nvidia cards have sharper outputs on my monitor compared to AMD cards. So there.

    Comment


    • #12
      Originally posted by zexelon View Post
      I must say I am a happy Nvidia proprietary driver user... I dont have to hack, cut, compile or beg Linux to run with it, it just works.
      no, it doesn't. it only works to the extend allowed by novideo for its slaves. my driver works without limitations and "I dont have to hack, cut, compile or beg Linux to run with it"
      Originally posted by zexelon View Post
      As an added bonus I also get hardware that tends to kick AMD in the teeth.
      another popular misconception among nvidiots

      Comment


      • #13
        Transitioned from a Nvidia GPU to a RDNA based one. I must say in the early days i had a lot of blackscreens and GPU crashes, but meanwhile the bugs are ironed out and it works like a charm, multimonitor, unplug / replug, standby, 10bit output, HDR / Deep Color, 165Hz, FreeSync everything just works!

        lack of good GPGPU support for neural learning on RDNA/RDNA2 is a bummer though, but meanwhile OpenCL support got quite good with rocm 4.0.
        AMD decided to split ways and have a compute focused (CDNA) and a graphics/rendering focused (RDNA) architecture and RDNA/RDNA2 isnĀ“t officially supported in their GPGPU stack (ROCM).. Some Kernels for deep learning are missing for RDNA / RDNA2 anyway.

        Comment

        Working...
        X