Results 1 to 9 of 9

Thread: Driving Linux-based Benchmarking With Sandtorg

  1. #1
    Join Date
    Jan 2007
    Posts
    14,904

    Default Driving Linux-based Benchmarking With Sandtorg

    Phoronix: Driving Linux-based Benchmarking With Sandtorg

    We have invested a lot of resources into enriching the Linux hardware experience particularly by improving Linux performance benchmarks and taking the necessary steps to make Linux-based benchmarking an attractive offer for hardware and software vendors. We have also strived to ensure that open-source developers understand the importance of automated testing and that they have the proper tools to fully automate tests relevant to them when looking for performance regressions and other conditions that otherwise would not easily be caught in an efficient and effective manner. At the same time, we have sought to standardize the benchmarking process of Linux desktops to make it easier for end-users and companies when looking to gauge how well something works on Linux. The Phoronix Test Suite has made immense progress over the past year, but today it is now time to expose our latest endeavor, Sandtorg.

    http://www.phoronix.com/vr.php?view=13695

  2. #2
    Join Date
    Dec 2008
    Posts
    160

    Default

    A lot of interesting progress on many fronts Michael!

    (Definitely think that maximizing automated testing/analysis on all fronts [performance, security, resources, code correctness] is a must for FOSS to be as efficient as possible with the developer/testing resources we have)

  3. #3
    Join Date
    Dec 2008
    Posts
    5

    Default

    it would be nice if the generated graphs offered at least more than 2 colors... dark green on green is stupidly pitiful (at least the ones phoronix uses on its website.) Are those two ugly colors the only ones available??

  4. #4

    Default

    Quote Originally Posted by soros View Post
    it would be nice if the generated graphs offered at least more than 2 colors... dark green on green is stupidly pitiful (at least the ones phoronix uses on its website.) Are those two ugly colors the only ones available??
    Uhhh it's completely configurable... From the GUI there's even a color wheel you can pick from with the preferences. If doing it from the CLI, there is an XML file that contains all of the color codes.

  5. #5
    Join Date
    Jun 2008
    Posts
    86

    Default

    Quote Originally Posted by Michael View Post
    Uhhh it's completely configurable... From the GUI there's even a color wheel you can pick from with the preferences. If doing it from the CLI, there is an XML file that contains all of the color codes.
    Perhaps the graphs you use in your benchmarking articles should be color coded. It's your job to show off what PTS can do, after all!

  6. #6
    Join Date
    Dec 2008
    Posts
    160

    Default

    "stupidly pitiful" / "it's your job" ...as the web is open to mis-interpretations, a little considered commentary might be received better...

    I for one thank Michael for his efforts as testing overlord ;-)

    Sounds like some people would like a different colour scheme or legend to make graphs easier to read. I know I wouldn't mind seeing a benchmark system shown with reviews (say having a core 2 along side a reviewed i7), should a system be available. Perhaps the test suite could include some "standard" scores (although that would likely become irrelevant or inaccurate very quickly)... that being said, I've largely just enjoyed the reviews and haven't yet played with the test suite yet to understand its capabilities (and on my old system I'm a bit scared to see the scores ;-) )
    Last edited by Craig73; 04-14-2009 at 09:42 AM.

  7. #7
    Join Date
    May 2008
    Location
    Parish, NY
    Posts
    159

    Default

    so... still no KDE/Qt benchmarks?

  8. #8

    Default

    Quote Originally Posted by jeffro-tull View Post
    so... still no KDE/Qt benchmarks?
    In what regard?

    gtkperf and qgears2 have long been in PTS that use GTK and Qt4, respectively, for testing.

  9. #9
    Join Date
    May 2008
    Location
    Parish, NY
    Posts
    159

    Default

    sorry, that was supposed to be in this thread.

    Anyway, I just figured Qt/KDE benchmarks didn't exist since they weren't used at all in the comparion between Kubuntu and PC-BSD

Posting Permissions

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