thelinuxlink.net

Linux vs. Windows Power Usage

http://www.phoronix.com/scan.php?page=a ... =880&num=1

When you take into account the fact that more and more old PCs get revived with Linux instead of being thrown in the dumpster, I think Linux is far more green. Linux geeks tend to keep their old machines around for years and years, the average Windows user buys a new PC every 3 years and chucks the old one in the rubbish. That seems a lot more wasteful than a few extra watts of power consumption.

Green aside, it'd be awfully nice if laptop tools and its ilk worked better. My battery consumption when running linux is just freaking *crazy* when compared to other people running similar or better hardware configs using other OSes.

Snarkout wrote:Green aside, it'd be awfully nice if laptop tools and its ilk worked better. My battery consumption when running linux is just freaking *crazy* when compared to other people running similar or better hardware configs using other OSes.

Tickless kernel is supposed to really help with laptop power usage. I don't know if anyone is shipping it yet.

I think the fact that this test was done on a laptop is pretty significant. Also, if you read the whole article, Fedora used the least power at idle.

Wally Balljacker wrote:When you take into account the fact that more and more old PCs get revived with Linux instead of being thrown in the dumpster, I think Linux is far more green. Linux geeks tend to keep their old machines around for years and years, the average Windows user buys a new PC every 3 years and chucks the old one in the rubbish. That seems a lot more wasteful than a few extra watts of power consumption.

I think that's a myth , linux users like to buy new hardware just as any windows user.
Sure they might keep there older computers longer.

Once the ALPM patches are merged into the kernel I think we will be in the lead (http://www.lesswatts.org/results/desktop/index.php)

Also this is just Ubuntu's default configuration, you can probably tweak some settings and get it down with the help of powertop. I'm sure the desktop search is taking a fair amount of that power.

allix wrote:

Wally Balljacker wrote:When you take into account the fact that more and more old PCs get revived with Linux instead of being thrown in the dumpster, I think Linux is far more green. Linux geeks tend to keep their old machines around for years and years, the average Windows user buys a new PC every 3 years and chucks the old one in the rubbish. That seems a lot more wasteful than a few extra watts of power consumption.

I think that's a myth , linux users like to buy new hardware just as any windows user.
Sure they might keep there older computers longer.

In the past couple of years, I have yet to see any of the linux users around here chunk a machine unless it was an eight or early 16 bitter. I have seen linux users pass older systems to other family members or charity where the systems would still be used. I have recently collected at least ten chunked machines which includes two laptops. None of them were from linux users. I will admit that hardcore linux (especially gamers) users who can afford it will get a new machine at the drop of a hat.

Gomer_X wrote:

Snarkout wrote:Green aside, it'd be awfully nice if laptop tools and its ilk worked better. My battery consumption when running linux is just freaking *crazy* when compared to other people running similar or better hardware configs using other OSes.

Tickless kernel is supposed to really help with laptop power usage. I don't know if anyone is shipping it yet.

Fedora 7 included tick-less kernel support for x86, Fedora 8 will have tick-less for both x86 and AMD64.

Tsuroerusu wrote:

Gomer_X wrote:Tickless kernel is supposed to really help with laptop power usage. I don't know if anyone is shipping it yet.

Fedora 7 included tick-less kernel support for x86, Fedora 8 will have tick-less for both x86 and AMD64.

The test in the article was done on Fedora 8 Test 3 (it says Fedora 7, but the caption says 8Test3). Maybe they didn't have tickless support enabled? My experience with test releases is that things can be pretty strange. Again, I don't think this is a very thorough and fair test.