Announcement

Collapse
No announcement yet.

Mesa 20.1-dev RADV vs. RADV+ACO vs. AMDVLK vs. AMDGPU-PRO Vulkan Radeon Linux Gaming Performance

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

  • mr_raider
    replied
    Sorry for resurrecting and old thread, but I just installed an rx5700 and want to make sure I got all the pieces right.

    I'm using KDE neon which is an 18.04.4 base. I'm using the stock 5.3 kernel, and upgraded the following options:

    1. Mesa 20.1 via kisak ppa

    2. Open CL from the amdgpu-pro package (OpenCL headless option)

    So far my native Linux games work fine, with one notable exception (Divinity original Sin), and I've got Folding@home chugging along.

    My next question comes with respect to a few windows native games I play either via Lutris (DXVK) or Steam Proton. My understanding is that Mesa has a "built in" vulkan driver called RADV and that should work. Also the amdgpu-pro package offers a vulkan driver I can install.

    For now would people recommend I leave my install as is, or should I install the vulkan components from amdgpu-pro. I'm not upgrading to a 20.04 based install until July at least, but I can manually upgarde the kernel if need be. I know 5.4 is supposed to be a big improvement.

    Leave a comment:


  • humbug
    replied
    Very informative! Also is it just my browser or is the text on the graphs borderline unreadable?

    Regarding ACO too I was thinking Valve has so much investment into MESA, into ACO, into the Vulkan initiative, into their Source 2 engine's Linux's and Vulkan support, into Steam Linux etc, porting their entire back catalog to Linux etc.... yet their big game release Half Life Alyx is not launching on Linux... go figure.

    Leave a comment:


  • akuhtr
    replied
    I know, its additional work, but would be nice to have windows as reference sample. Games are developed/ported for windows as main target platform. Right now, we can compare linux drivers, but we don't know what is state of the art for this games.

    Leave a comment:


  • user1
    replied
    Wow, I'm really surprised by the performance of AMDVLK. I do suspect however that it is now generally faster than AMDGPU-PRO because it's the latest version which was released just a few days ago, and the latest AMDGPU-PRO is 3 months old.

    Leave a comment:


  • loganj
    replied
    thank god there are so many drivers for you (amd users) to pick one

    Leave a comment:


  • smitty3268
    replied
    Hmm,

    The -pro drivers really don't have any performance advantage anymore, over either the amdvlk or radv drivers.

    The amdvlk performance looks pretty good, but the strange brigade results shows AMD isn't taking it seriously as a gaming driver yet, just a place to dump open sourced code. That behavior is most likely due to a poor implementation of vulkan's mailbox presentation mode - which is integrated with X11, and therefore likely written for scratch in this driver and not copied from something else.

    And the Vega 56 issue in ACO is rather unfortunate for this set of tests, though it sounds like it's already been fixed up.

    Leave a comment:


  • tildearrow
    replied
    Typo:

    Originally posted by phoronix View Post
    this appears to be a regression in the AMD stack with the AMDGPU-PRO 19.50 perfoming okay and it being an older revision.

    Leave a comment:


  • microcode
    replied
    Originally posted by Venemo View Post
    Sorry guys, the Vega problem was my fault. I don't have a Vega myself so didn't have the opportunity to test ACO tessellation there... Now that the problem has been fixed, it would be nice if Michael could repeat the tests with those fixes included.
    I think I have an extra (Vega 64) I could send you when Corona-chan moves on.

    Leave a comment:


  • Venemo
    replied
    Sorry guys, the Vega problem was my fault. I don't have a Vega myself so didn't have the opportunity to test ACO tessellation there... Now that the problem has been fixed, it would be nice if Michael could repeat the tests with those fixes included.

    Leave a comment:


  • qarium
    replied
    yes amd and the FLOSS community really did the job.

    now the biggest problem is the legacy stuff of outdated CAD software using non-standard OpenGL
    if we can manage to solve this problem then AMD can just turn of the plugs from the old closed source driver.

    this PRO/FireGL customers according to Bridgman do cross-subsidizatizing the linux/FLOSS development.
    But if money isn't the problem of this Rich customers why they do not use the money to fix their software?
    or just write profiles for mesa to run the Legacy apps.

    however i think geting rid of the Legacy closed source driver should result in good PR dor AMD.

    Leave a comment:

Working...
X