New Ubuntu eats more power than Windows 7

As someone who has done some testing with Windows, Ubuntu, and other distributions I have to side with the author on this one.

Ubuntu does not do a good job of managing battery life. The new "power" stuff in 10.04 tunes 3 kernel parameters, just look at the package source. Fewt was doing that *2 years ago* in his Eee PC Tray software I use on my Eee. I tested 10.04, and went back to Eeebuntu because I get less than 1/2 the battery life that I did with Eeebuntu plus 10.04 for some reason is not very stable anyway. I dual boot to Windows 7 on this thing too and let me tell you that Ubuntu 10.04 just does not compare in terms of battery life. I have been testing the new Eeebuntu called EB4 and it has this thing called "Jupiter" built in that is based on Eee Tray. I love it.

I own a ASUS Eee PC 1201n, and this article is bullshit . . . I am sure any windows fanboy will enjoy reading it, and thinking their OS is finally up to par, but the truth is it is not. When my computer arrived I said, I will try Win7 for one week, maybe even dual boot, maybe but problem not convert to Win7 full time. BUT WHEN I TURNED IT ON WINDOWS WAS UNUSABLE. I was sad and thought I wasted a bunch of money, but when I loaded Ubuntu on it, Lucid Lynx, it ran fast!!!!!! Now I am on 64bit, and that does help the dual core Atom processor. I tried 32bit Ubuntu also, and it was slower but still workable unlike Windows. But hey, everybody knows the 1201n has bad battery time already because of the graphics card, so who cares! Oh and I understand why it was only a idle test, because Windows on this computer can only do that or it fucking explodes.

This article gives totally an inaccurate picture as some commentators above mention. There are so many holes in this that I don't know what point you are trying to make, actually maybe I do.

I am using Ubuntu 10.04 Netbook edition on my Dell Latitude 2100, and it installed easily, and everything, yes, worked out of the box. It looks and performs fantastic. Just as Windows7 has problems, Ubuntu 10.04 has a few as well although I haven't noticed any yet. As a comment above says, Ubuntu 10.04 is the closest they have come so, I thank many thousands around who have devoted their spare time so that I can use and enjoy a hassle-free OS.

Windows7 is made by one of the most profitable, powerful and influential corporations in the world. I expect it to be the easiest, and absolutely hassle-free OS on the planet, however, it is not and doesn't even come close. Microsoft works very closely with OEMS's, not only to tweak Windows7 so that it is optimised for the hardware but also to keep Linux OS's out of their lineup. Although some do such as Dell, it is hidden, not easily visible, certainly not on retail floors, and their staff don't even know they have Linux PC's in their lineup.

As for the power comparison, it compares very favourably to a Windows7 netbook, so that I cannot make a call either way since there are many variables involved. I know because I compare it with mine. My result is different than this article.

As another comment mentioned above, I installed Ubuntu 10.04 myself without no tweaking and customisation for the hardware at all, the total opposite with a Windows7 OEM install.

Even if I got 5% less battery time, using Ubuntu 10.04 it is a no-brainer to me. Windows will suck your time out of your personal life, I have been there, and not going back, because I have been exposed to better alternatives. Also most people I know have purchased new PC's just because the PC's have slowed down over time because of using Windows. I have been running Ubuntu 8.04 on my 4yr old Dell for 2 years and it has the same performance as when I first installed it. Linux OS's are ready for the masses.

Does Microsoft really need you to advertise Windows7 for them and well as putting down any better alternatives? I don't think so. Microsoft already has more than enough funds and foot-soldiers to do that work.

I bought a wonderful little Samsung NC10 netbook around the middle of last year.

Used it for a while, figured that since I wasn't going to do anything 'serious' on it, just e-mail & web, I could get away with giving Ubuntu a try.

Noticed the decreased battery life, slideshow flash video playback and instantly reinstalled Windows. Never looked back. Never going to either, although I do occasionally boot Fedora off a USB stick for reasons of wireless (in)security.

It just shows that the "Drop this OS inside any PC" concept that Microsoft pioneered is entirely flawed.

Either you like it or not, each PC is different.
So, the idea of an OS being magically able to run everywhere, regardless of underlying hardware and peripherials, is simply impossible to achieve. Take the ASUS for example: I bet 10 bucks that there is some hardware inside that netbook that somehow Ubuntu didn't see properly. Instant failure.

Sure, Windows can run on pretty much any hardware, but only after a bunch of drivers "optimizes" (read: screws up) the operating system.

Am I the only one that sees the need to standardize the peripherals and hardware of the PC world?
Why, oh why, should there be a hundred different drivers for sound cards, when there is only one for cardreaders and DVD drives?

Failing that, I think we really need some laptops with the "Designed for Ubuntu" sticker on them.

If you can't standardize the peripherals and hardware, at least you need to make sure the laptops you sell have standardized hardware in them.

1) Measuring AC input power has no necessary relationship to power consumption when on battery power. Measuring the latter is *quite* tricky (to make physical connections) unless the unit has built-in provisions to do so. Also, all power conversion is lossy and has some overhead too. -- SO what's the point of measuring *AC* input?

2) Windows 7 has a definite problem with battery life on some systems, even from new. My guess as to possible reason includes that it (or BIOS) wrongly measures the battery as full and isn't charging it. That view could fit the observations. -- Conversely, Ubuntu may well be *properly* charging the battery, but "losing" this contest!

3) IF Windows 7 isn't working correctly on the test system by not charging the battery fully, it'd gradually sap the apparent battery capacity. A long term test is needed.

4) Just for a fillip, but could be highly significant: what's the maximum crest factor for the RMS converter in the given meter? Is it at *any* time accurate for rectified loads? Perhaps *you* believe the apparently unambiguous number that a digital meter shows, but I have some small notion of the complexities involved.

Taking only item 1 as not in dispute: the stated method is nearly useless. Adding in 2 and 3, it may well be an entirely wrong conclusion.

This just goes to show the downside of a free product. It might be a good product, which ubuntu is, but since it's free odds are there isn't as much research and testing going into it accross all hardware platforms before it launches so you'll run into quarks like this. That said now that the problem is known expect them to work on it regardless of the special netbook version.

@marth
Actually windows7.com is completely worthless because of internationally recognized copywrite law pass a few years back. Basically the law stipulates that any web address which could logically be related to a given copywrite/trademark belongs to the holder of that copywrite. In otherwords someone could drop a billion for the url then have MS come along 10 seconds later saying "give it to us" and they would be required to by law for absolutely free.

Well that's it! I'm switching to Vista as of right now. I know power consumption was the reason I left Windows behind forever, and so it makes sense to go to Vista now.

Thank you team for bringing this issue to the deluded Linux users. Please continue with your good job.

First, your numbers don't add up, you should fix the typos

Second: A question. By "out of the box" what do you mean?

Are you comparing a computer that came from the OEM with Windows with an identical piece of hardware that ALSO came from the OEM with Linux? Or did you install the Linux?

The way an operating system is set up on a piece off hardware determines how the power saving/management works (if at all) If you did what I think you did, then you need to redo it.

Otherwise, my apologies for guessing that you did it wrong.

I have to wonder what goes through someone's mind when they make patently false statements about someone or something. They must lead pretty miserable lives all things considered. Have to if they can feel willing to make stupid claims such as the anonymous coward did about Ubuntu Lucid 10.04.

This version of Ubuntu though has some problems is the closest they have come to perfection and it really does put it in a category to challenge Win7 and Mac OSX.

Every 6 month release is going to have issues and every OS has them with every update for someone. But to purposefully go to public arenas and make patently false claims is akin to a tad bit of insanity on that person's life, and a lack of self control.

Given the sheer number of things people do with their computer and given how well an OS has to be written to accommodate a portion of those things, one has to give tremendous praise to the Linux community and Canonical for making such a near perfect release.

I'd say to the Anonymous Coward. Grow up and think like a man, a real man. Stop acting the child. Put your prejudices behind you and MOVE ON.

Ubuntu comes with a netbook remix version and it means "install netbook remix on netbooks!".

what you don't understand ?
The setup described at Phoronix's article at the very beginning. Just click the link. And the method was measuring "AC power consumption of the netbook and notebook using a SeaSonic Power Meter".
AFAIK, both OSes change power management profile when you disconnect the AC power. But it's much harder to measure.

Numerous reports of a fraction for battery versus XP.
http://blogs.msdn.com/e7/
M$ hasn't commented on it there since February.

In any case, I'd like to see the exact test setup used to measure power consumption -- didn't find it on site -- but a simplistic method could easily render this data irrelevant.

Installed Ubuntu 10.04 on AMD platform.

cat /proc/cpuinfo shows that while system is idle all cores are clocked at nominal speed.

apt-get install powernowd

with AMD CPU frequency driver (powernow) installed all cores clock down to 800MHz while system idle.

So yes, Ubuntu 10.04 does crappy job by ignoring installation of CPU frequency driver, at least on AMD systems.

That may contribute to power consumption.

just my 2c

The power savings with DEFAULT are indeed in favor of Windows.
From my personal experience, Ubuntu (and most other distros) uses ondemand CPU frequency governor. This results in quicker responses. Why Ubuntu developers decided to use 100% brightness on battery, i don't know.
Anyway, Windows 7 is the best OS from MS ever. I still prefer Linux since i know what I want and know how to do it. Not to mention to keep my settings regardless Distro, ISA (x86, x86_64, ARM etc) or just an upgrade (SW or HW).

I have been running Lucid on a new dell studio hybrid and a dell mini 9. The power usage is almost identical. The main difference is Lucid Lynx is so much faster opening files like pictures music etc. I see no difference on the mini 9 with windows 7 and Lucid Lynx. For me personal Ubuntu wins hands down in performance and security. Plus I love the new interface.
Even if it used more power( I am taking that with a grain of salt) I am sticking with Lucid Lynx. My studio hybrid wireless keyboard multimedia keys work perfect as does everything else.Thank You Linux people!

The article said, "Ubuntu "Lucid Lynx" went through less power, but Windows still did better and the difference turned into a 65 per cent."

Anyone care to explain how it went through less power but lost another 9 percent to Windows. Does this sound fishy to anyone else.

Its all about the tweaking and the drivers and there is an official name for it - Wintel. Do we need to say more ? No need to kill the competition and have the regulators on your back - just make sure to find the right "friends" who will guarantee the competition will always be a bit "behind" on the friends' hw. So simple.

I am waiting for the ARM smartbooks - then it will become much more interesting. Sooo f*ckin' tired of this oligopoly in IT. I'm working with the biggest - exactly in the hw-enablement/drivers area.You can't even imagine..

Agreed, as long as it's using MeeGo, intel will guarantee you will get longer battery life. Ubuntu people will not be told the secret, that is why Intel ditch them in MeeGo's development after Ubuntu under delivered its moblin to Intel

We have several Linuces on a Hell notebook.

It just so happens that our UAT process included testing idle power consumption. We measured the idle power consumption *at the wall* with the lid closed, on a fully-charged battery, on mains power, using all default drivers.

Long story short: Windows XP: 11 watts; Ubuntu Server 10.04: 11 watts.

OK, so thats XP and not Windows 7. Then again, feh.

There are plenty of reasons to hate Ubuntu 10.04 - the fact that it utterly, utterly sucks compared to 8.04, for example. But idle power consumption isnt one of them.

Then again... these ARE the statistics for idling. What about the computer's performance when you're actually DOING something.

But still... it hurts. Power management matters suddenly. Who knew? :-)

Also, does this Ubuntu edition have a "Netbook Remix" yet? I think that's a semi-official spin-off that gets more out of your mini-microcomputer. And, maybe, is the place to fix this, as far as possible and needed.

Also, on a "Long Term Support" edition, you should expect to take time to set the best configuration - you'll probably be living with it for a while. Of course that's true of Windows 7 as well. "Palin" is still a long way in the future - in computer industry time.

windows7.com is worth $1,872!

http://www.statzeus.com/www.windows7.com