Announcement

Collapse
No announcement yet.

Intel HD Sandy/Ivy Bridge Linux Graphics Performance On Ubuntu 13.10

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

  • Intel HD Sandy/Ivy Bridge Linux Graphics Performance On Ubuntu 13.10

    Phoronix: Intel HD Sandy/Ivy Bridge Linux Graphics Performance On Ubuntu 13.10

    With most of the exciting Intel Mesa open-source Linux graphics driver updates being now about Haswell or Bay Trail (Valley View), it's been a while since last publishing any benchmarks of previous-generation Intel "Sandy Bridge" and "Ivy Bridge" graphics results with the latest open-source driver code. Changing that, today are benchmarks of several different Intel Sandy Bridge and Ivy Bridge processors when running Ubuntu 13.10 with the Linux 3.11 kernel and Mesa 9.2.

    Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite

  • #2
    What about some AMD comparisons, like Kabini? For example, this looks interesting.

    Comment


    • #3
      Originally posted by Cyborg16 View Post
      What about some AMD comparisons, like Kabini? For example, this looks interesting.
      I don't have many AMD setups at all aside from those I purchase. I only have a couple APUs at all. From Intel it's easy to get hardware where as of lately from AMD it's basically been impossible.
      Michael Larabel
      https://www.michaellarabel.com/

      Comment


      • #4
        Browser bench

        Hey Michael

        How about adding some browser benchmarks? Maybe shipping V8 and some js benches for CPU. And use some lib with webkit for rendering benchmarks. Or just packaging node.js with some available benchmarks.

        Comment


        • #5
          Originally posted by talvik View Post
          Hey Michael

          How about adding some browser benchmarks? Maybe shipping V8 and some js benches for CPU. And use some lib with webkit for rendering benchmarks. Or just packaging node.js with some available benchmarks.
          So far I haven't found a good way to automate it completely.. There's been a few different projects for trying to automate tasks within web browsers for unit testing, but nothing that I've been happy with.
          Michael Larabel
          https://www.michaellarabel.com/

          Comment


          • #6
            Originally posted by Michael View Post
            So far I haven't found a good way to automate it completely.. There's been a few different projects for trying to automate tasks within web browsers for unit testing, but nothing that I've been happy with.
            Instead of using the browsers, just go directly to the engines. v8, node.js, QtWebEngine... could be easy to package.

            One example, octane running on node.js:
            Octane benchmark for Node.js. Contribute to dai-shi/benchmark-octane development by creating an account on GitHub.

            Output:
            Code:
            % git clone https://github.com/dai-shi/benchmark-octane.git
            % cd benchmark-octane
            % node run.js
            Richards.Richards x 288 ops/sec ?10.28% (45 runs sampled)
            DeltaBlue.DeltaBlue x 186 ops/sec ?9.87% (49 runs sampled)
            Crypto.Encrypt x 155 ops/sec ?9.77% (48 runs sampled)
            Crypto.Decrypt x 7.99 ops/sec ?14.10% (26 runs sampled)

            Comment


            • #7
              Originally posted by Michael View Post
              So far I haven't found a good way to automate it completely.. There's been a few different projects for trying to automate tasks within web browsers for unit testing, but nothing that I've been happy with.
              I also like this solution seems more versatile, http://src.chromium.org/svn/trunk/sr...mark_runner.py

              It runs the browser with the benchmark address as argument, and then listens for the results to be posted from the benchmark.
              It only takes a couple of lines of python script, and requires maybe only one line change to the benchmark.

              Comment

              Working...
              X