Announcement

Collapse
No announcement yet.

Ubuntu 8.04 vs. Windows Vista Power Usage

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

  • Ubuntu 8.04 vs. Windows Vista Power Usage

    Phoronix: Ubuntu 8.04 vs. Windows Vista Power Usage

    In Q4'07 we had looked at Ubuntu's power consumption with all of their Linux releases going back to Ubuntu 5.04. While Linux has improved in recent years when it comes to power efficiency and optimizations, more processes running on the desktop had canceled out any real power improvements. Following that article was a look at power consumption between Windows and Linux. We had used an old desktop system in that comparison and Ubuntu 7.10 was consuming the most power while idling but Fedora 8 Test 3 had consumed the least amount of power and had beat out both Windows XP and Vista. While using the desktop, however, both versions of Windows had consumed less power than Fedora and Ubuntu. With Ubuntu 8.04 LTS now available, we have decided to run another simple power comparison. This time we are using a Lenovo ThinkPad notebook and an AMD server as we see whether Ubuntu Hardy Heron or Microsoft Windows Vista consumes less power.

    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
    The problem with vista is that by default it is set on balanced and not on power saving mode. In balanced with ac power cpu clocks are not set to reduce to their lowest levels, whereas in modst linux distro's ondemand clocks right down to the minimum. Balanced also does not shut down things like pcie powerlevels etc.

    An interesting test on laptops would be to see how long they take to discharge when on battery mode

    Comment


    • #3
      I've had vista with ubuntu 7.10 in dual boot prior installing 8.04 as my only OS. My laptop is HP nx9420. With the use of powertop and aticonfig utility, battery on both 7.10 and 8.04 lasts 3 hours at most, while on M$ Windows Vista Business, it lasted cca 4 hours in powersave mode (with aero off).
      Strange is that if I turn off compiz effects, I get cca 5 mins extra battery life, so I believe it's not worth turning compiz off in order to save your battery life.

      That's how I concluded that vista spends less power on my laptop, and I can explain that with the following statement: vista's power consumption is directly proportional to it's performance

      Comment


      • #4
        Michael, did you set Windows Vista to power saving mode? Did you have Aero switched on?

        Thanks for the article anyway, but I'd be interested to know how Fedora 9 compares to Ubuntu 8.04 and Vista. It would be a great if could test it, also.
        Last edited by pirast; 27 April 2008, 04:26 PM.

        Comment


        • #5
          nice little test, but it seems silly to test an opteron server on two desktop OS's. I mean where you running X on a server? and vista ultimate on a server ?? ahahaha surely you should try windowns server 2008 against centOS or something abit more serious... maybe the 8.04 server edition if you want to stick to ubuntu.

          Also you mention no 3rd party drivers? does that mean the windows box was driverless? Or just basic drivers?

          It seems more like a bit of fun than a serious test - which maybe it's supposed to be....

          and don't bother testing fedora 9 against ubuntu, the difference is probably nothing, and if there is one, it's probably just because of room temperature being slightly higher for one test than another.

          Comment


          • #6
            Hey Michael,
            Would you be able to do a power consumption test between 7.10 i386 and AMD64 vs 8.04 i386 and AMD64 to see if the difference between the two has been narrowed and if 8.04 is better than 7.10 for power?

            Thanks.

            Comment


            • #7
              and don't bother testing fedora 9 against ubuntu, the difference is probably nothing, and if there is one, it's probably just because of room temperature being slightly higher for one test than another."]and don't bother testing fedora 9 against ubuntu, the difference is probably nothing, and if there is one, it's probably just because of room temperature being slightly higher for one test than another.
              The Fedora devs have tried to improve power savings a lot in recent days, and they ship a newer kernel than Ubuntu does. Why not compare it?

              Comment


              • #8
                Despite the test, I have often notice that I get better battery performance in windows than in Linux. Despite the great gains over the years, I still find Linux to be probmatic on most laptops. There's always something that doesn't work and sleep/suspend are often not working properly. I have for example tried to get my Averatec 2370 working with Linux but have finally given up. Next time I buy a laptop, it will have to be one that's more Linux friendly.

                Out of curiousity, have anyone like a particular distro over another for laptops? Phoronix has pointed out that Fedora was getting (slightly) better battery life.

                Comment


                • #9
                  Well, it's a nice review, but I think for the test done on the Thinkpad more people are concerning the power consumption when the laptop is running on battery instead of AC.

                  Personally, on my own Thinkpad T61p, I don't see any significant improvement on battery life from Ubuntu 7.10 to 8.04 (both 64bit), even though it was said that power consumption has been reduced significantly from kernel 2.6.22 to 2.6.24 especially on the 64bit platform. When using laptop during travels, I still find myself boot into Windows XP, which provides at least one additional hour of battery life (3-3.5hr vs 2-2.5hr) comparing to Linux.

                  I agree comparing power consumption with different OS is not actually comparing apple with apple, but I think it still will be very useful to compare battery life during normal usage (wifi surfing with 50% LCD brightness, CPU profile set to be "conservative" or "power save") for most laptop users.

                  Comment

                  Working...
                  X