Page 1 of 2 12 LastLast
Results 1 to 10 of 87

Thread: The Most Comprehensive AMD Radeon Linux Graphics Comparison

Hybrid View

  1. #1
    Join Date
    Jan 2007
    Posts
    14,369

    Default The Most Comprehensive AMD Radeon Linux Graphics Comparison

    Phoronix: The Most Comprehensive AMD Radeon Linux Graphics Comparison

    Last month I alluded to a 40-way graphics card comparison being worked on at Phoronix. This comparison is to extensively compare the performance of the open and closed-source drivers for each graphics card and to comprehensively comment on other areas of the Linux graphics driver support. Not only is the OpenGL performance being evaluated, but the thermal performance, CPU utilization, and power consumption is being looked at too. Being published today to mark the beginning of the Oktoberfest 2011 articles are the ATI/AMD Radeon results. This includes 28 of the 40 graphics cards, with GPUs as old as the Radeon X800XL and as new as the AMD Radeon HD 6950.

    http://www.phoronix.com/vr.php?view=16438

  2. #2
    Join Date
    May 2007
    Posts
    319

    Default

    FFS you could at least try fact checking maybe you could go on a journalism night course instead of going to Oktoberfest next year

    The weather channel had nothing to do with the r300 classic driver. It was reverse engineered by a team of 5-7 people over years. I dont think anyone was funded for it ever. The WC funded r100 and r200 drivers which were nothing like the r300.

  3. #3
    Join Date
    Aug 2009
    Location
    Russe, Bulgaria
    Posts
    500

    Default

    Quote Originally Posted by airlied View Post
    FFS you could at least try fact checking maybe you could go on a journalism night course instead of going to Oktoberfest next year

    The weather channel had nothing to do with the r300 classic driver. It was reverse engineered by a team of 5-7 people over years. I dont think anyone was funded for it ever. The WC funded r100 and r200 drivers which were nothing like the r300.
    OMG, Dave! You have read the entire article??
    I just skimmed through it. It doesn't tell us anything new

  4. #4
    Join Date
    Apr 2011
    Posts
    35

    Default tl;dr

    Dude... an article with 7 billion pages and (as always) no conclusion which states the most important observations after the last test.

  5. #5
    Join Date
    Nov 2010
    Location
    Stockholm, Sweden
    Posts
    394

    Default

    Quote Originally Posted by mark_ View Post
    Dude... an article with 7 billion pages and (as always) no conclusion which states the most important observations after the last test.
    I read half of it, then started skimming. But one observation I quickly made (just skimming the graphs) was that there's to point in upgrading from my 4850 to a 5* or 6* since there' virtually no performance gain on the FOSS drivers. Unless I switch to Catalyst that is.

  6. #6
    Join Date
    Oct 2007
    Posts
    1,259

    Default

    Quote Originally Posted by mark_ View Post
    Dude... an article with 7 billion pages and (as always) no conclusion which states the most important observations after the last test.
    No table of contents and no conclusion = too long/didn't read/ABP still on. Oh, and someone give Marek an honorary membership (I think he's earned it).

  7. #7
    Join Date
    Dec 2007
    Location
    Edinburgh, Scotland
    Posts
    574

    Default

    It's not a comprehensive comparison if you don't have a conclusion. I'm sure these tests take a long long time to run on each card but it' pretty pointless wasting all that time if you don't follow through with decent analysis

    Since advertising and Phoronix Premium aren't paying you enough you might need to look into some form of Corporate sponsorship - that probably would mean cleaning up your language and fewer articles on beer though

  8. #8
    Join Date
    Jan 2009
    Posts
    614

    Default

    I have found a few lies and inconsistencies in the article.

    1) VDrift doesn't work, because it has non-compliant shaders that shouldn't compile according to the OpenGL standard. They can be fixed easily, though.

    2) Unigine Heaven, Unigine Tropics, and Unigine Sanctuary work correctly and beautifully with floating-point and S3TC textures with Mesa on both r300g and r600g.

  9. #9
    Join Date
    Nov 2010
    Location
    Stockholm, Sweden
    Posts
    394

    Default

    Quote Originally Posted by marek View Post
    I have found a few lies and inconsistencies in the article.

    1) VDrift doesn't work, because it has non-compliant shaders that shouldn't compile according to the OpenGL standard. They can be fixed easily, though.

    2) Unigine Heaven, Unigine Tropics, and Unigine Sanctuary work correctly and beautifully with floating-point and S3TC textures with Mesa on both r300g and r600g.
    Hmm, point 1 sounds interesting, perhaps I should have a look at that when I get home =)

  10. #10
    Join Date
    Jul 2010
    Location
    Melbourne, Australia
    Posts
    115

    Default Im glad

    Haha Im actually glad that I bought a Phoronix Premium membership to read this article

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •