Page 2 of 6 FirstFirst 1234 ... LastLast
Results 11 to 20 of 59

Thread: Intel vs. AMD Performance-Per-Watt On Ubuntu 14.04 Linux

  1. #11
    Join Date
    Jan 2011
    Posts
    360

    Default

    Quote Originally Posted by devius View Post
    These results are all wrong again.

    Let's take the case of the Timed Kernel Compilation with the i3-4130 and the A10-7850. They both take about the same time to finish the test, but the i3 uses 40W less power on average to do so! That's about 34% less used by the i3 to do the same job in the same amount of time, which means the i3 is more power efficient than the A10. However on the bogus "performance-per-watt" chart the A10 appears with a much better score.
    Indeed, the i3 4130 and the A10 take the same time, the A10 needs 120W almost the whole time but the i3 needs 75W, so it’s obvious the i3 has better performance per watt. Yet the perf per watt graph shows 0.71 (what unit is this btw?) for the A10 and 0.47 for the i3.

    Apart from that big problem, it’s an interesting article…

  2. #12
    Join Date
    Jan 2010
    Location
    Somewhere in Kansas.
    Posts
    257

    Default

    Actually, these results impress me with amd.

    On the CPU test the A10 was in between the i5-3470 and the i5-4670.
    the i5-3470 costs $190 (on Newegg)
    the A10-7850K costs $190.

    So, if we compare the two that cost the same, the A10 gives you better CPU performance per watt and better graphics.

    Other prices:
    i7-4770K $340
    i7-3770K $330
    i5-4670 $210

  3. #13
    Join Date
    Jan 2010
    Location
    Somewhere in Kansas.
    Posts
    257

    Default

    Huh. Unless the other posters are right and the charts have something wrong with them. I'm gonna do some math here quick.

  4. #14
    Join Date
    Jan 2010
    Location
    Somewhere in Kansas.
    Posts
    257

    Default BUG phoronix test suite.

    Ok the formula the power efficiency chart seems to use is this:
    Watts / time
    (using this gives me similar results to what is written in the article.)

    which means that the longer the computation took, the smaller the number. The formula used for the charts is backwards.

    It needs to be time/watts.

  5. #15
    Join Date
    Sep 2013
    Posts
    24

    Default

    I'd like to see total work done over total energy used numbers.

  6. #16
    Join Date
    Sep 2013
    Posts
    24

    Default

    Quote Originally Posted by vick View Post
    I'd like to see total work done over total energy used numbers.
    Scratch that. I want total work / total time / total energy.

  7. #17
    Join Date
    Jan 2010
    Location
    Somewhere in Kansas.
    Posts
    257

    Default

    http://phoronix.com/forums/showthrea...d=1#post390102

    Yeah. Just posted a post in the test suite forum. Hopefully Micheal will see it or this forum thread.

    Forumlas that would work.
    Watts x time = joules (Smaller number would be better)
    time/watts (smaller number would be better)

  8. #18
    Join Date
    Jan 2010
    Location
    Somewhere in Kansas.
    Posts
    257

    Default

    Quote Originally Posted by vick View Post
    I'd like to see total work done over total energy used numbers.
    The workload was the same.

    So you would just need to know the joules used.

    Watt x time = joules.

  9. #19
    Join Date
    Jan 2014
    Posts
    2

    Default

    Why did you use the ondemand governor for the AMD cpu whereas the intel processors all have the performance governor set?

  10. #20
    Join Date
    Jan 2011
    Posts
    360

    Default

    Quote Originally Posted by ua=42 View Post
    time/watts (smaller number would be better)
    No, performance per watt would be 1/time/watt (or speed/watt if you prefer).

Posting Permissions

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