Page 9 of 9 FirstFirst ... 789
Results 81 to 87 of 87

Thread: The Most Comprehensive AMD Radeon Linux Graphics Comparison

  1. #81
    Join Date
    Dec 2007
    Location
    Edinburgh, Scotland
    Posts
    574

    Default

    Quote Originally Posted by log0 View Post
    Interesting comparison(I mean the thread topic). I find the CPU load numbers quite puzzling, 3-5 times lover than catalyst. http://www.phoronix.com/scan.php?pag...berfest&num=36

    I somehow had the idea the open source drivers would be CPU limited. No multithreading with catalyst too, load < 25% for a 4 core CPU? I asume 100% is full CPU utilisation. Would be great if one of the driver developers could comment on this.


    To the Phoronix and Michael critics: Stop derailing the thread! Want to whine about it? Get you a fscking blog!
    I think you might be right. Single threaded apps will have a lower cpu utilisation - would be great if we could see the utilisation of each core during the test

  2. #82
    Join Date
    Dec 2007
    Location
    Edinburgh, Scotland
    Posts
    574

    Default

    Quote Originally Posted by jrch2k8 View Post
    about the whiners i dare you to point me out another site which at least has 10% of the information richness about linux and alikes
    beside phoronix (The H maybe but most articles about OSS are from phoronix and linux.com follow the same direction)
    You'll find http://lists.freedesktop.org/mailman/listinfo is where most of the phoronix articles are from along with http://planet.freedesktop.org/

    If you use RSS aggregation and subsribe to some planets like freedesktop or fedora and subscribe to some of the above mailing lists (filtered in gmail) you'll find you get all the info of this site

    Michael doesn't pull these articles out of his ass (even though sometimes they may read like they have)
    Last edited by FireBurn; 09-23-2011 at 09:48 AM. Reason: extra info

  3. #83
    Join Date
    Feb 2008
    Location
    Linuxland
    Posts
    4,729

    Default

    While following the lists does get you the same info, the volume is rather huge. I appreciate how Michael filters the important bits for us.

  4. #84

    Default

    Quote Originally Posted by log0 View Post
    To the Phoronix and Michael critics: Stop derailing the thread! Want to whine about it? Get you a fscking blog!
    Michael himself derailed the thread, by whining about adblock in his article (twice).

  5. #85
    Join Date
    Jul 2010
    Posts
    373

    Default

    Quote Originally Posted by AnonymousCoward View Post
    Michael himself derailed the thread, by whining about adblock in his article (twice).
    And for the majority of the posters this seems to be the only info from a 38 pages article worth to be discussed. Sorry but I don't get it.

  6. #86
    Join Date
    May 2011
    Posts
    353

    Default

    This is not the only article where Michael mentions AdBlock. So let's open a seperate topic for it and use this one for discussions about Radeon cards on Linux.

  7. #87
    Join Date
    Oct 2008
    Posts
    2,908

    Default

    Quote Originally Posted by log0 View Post
    Interesting comparison(I mean the thread topic). I find the CPU load numbers quite puzzling, 3-5 times lover than catalyst. http://www.phoronix.com/scan.php?pag...berfest&num=36

    I somehow had the idea the open source drivers would be CPU limited. No multithreading with catalyst too, load < 25% for a 4 core CPU? I asume 100% is full CPU utilisation. Would be great if one of the driver developers could comment on this.


    To the Phoronix and Michael critics: Stop derailing the thread! Want to whine about it? Get you a fscking blog!
    I think the issue with the OSS drivers isn't doing lots of math or other calculations on the CPUs. It's them sending/fetching data to the GPU and then sitting there waiting on it before it can do anything else. That ends up showing the CPU as idle instead of busy, but it's waiting and can't do any more work.

    I expect the higher CPU usage with Catalyst is mostly due to just pushing a lot more data to the GPU because of the higher performance, but it's also possible that it does more work optimizing everything in the driver and that ends up costing it more than the simpler OSS driver.

Posting Permissions

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