Is it possible for a HW agnostic OS like Linux to beat a HW specific OS when it comes to Power Consumption. I assume that Linux would require quite some tweaking to perform the same.
Phoronix: Fedora 17 Doesn't Change The Apple MacBook Pro
Following yesterday's OS X 10.8 Mountain Lion vs. Ubuntu Linux benchmarks and the OS X vs. Linux power consumption results after that, some wondered whether Ubuntu was to blame for the poor Linux showing on the Apple hardware. Unfortunately, Ubuntu isn't alone and here's some fresh data from Fedora 17 on the MacBook Pro...
http://www.phoronix.com/vr.php?view=MTE2NzQ
Is it possible for a HW agnostic OS like Linux to beat a HW specific OS when it comes to Power Consumption. I assume that Linux would require quite some tweaking to perform the same.
Michael, could you rerun the power tests under Fedora 17 KDE? Dont have to post an article about it if you dont want, but just a link to the results would be nice. I ask because Compiz/Unity have been refactored like crazy in the recent year or so, Gnome (mutter) got a major makeover with the 3.x, im curious to see what Kwin is doing since that codebase has been in the "optimization" stage of development for a couple of years now, plus Martin has even made a few blog posts directly related to Kwin/KDE power management. Atleast this would help in regards to narrowing down what is causing the increased power usage, whether its terminal-userspace (background processes) vs graphical userspace (Gnome/KDE/Unity) vs all in the kernel.
It doesn't make sense at all if the author used opensource graphics driver for testing. It's well known that opensource drivers suck at power management.
I'm surprised that Michael hasn't done his research on this.
Obviously the power savings on a Macbook Pro will come from using the on-board Intel HD graphics, and disabling the discrete graphics. To do this, you need to:
1. Boot in EFI mode
2. Switch GMUX so that the Intel card will output to the screen
3. Disable discrete GPU (ATI/NVidia depending on model)
It's not bad with kernel 3.5, but earlier kernels required patches to make this work properly.
I can now achieve 12W idling on my 2011 Macbook Pro and about 5 hrs battery life!
Absolutely it's very possible despite what the appleites and consolers might tell you. The main issue is drivers, which unless someone says otherwise I'm going to assume that the MBP 2010 was their usual Intel IGP + Nvidia gpu, which means yeah... no optimus yet and nouveau not really being set up for power management yet, of course it's going to be more power hungry. The other issue is potentially undocumented EFI hacks that Apple might be using. Since they do control the entire stack it is very possible for them to have EFI and thus ACPI weirdness in an attempt to cut out people from running other OSes on their systems, or at the very least to cast them in a worse light.
This said let's see KDE and also Windows results out of curiosity.
Why is Windows 8 not part of the upcoming comparison? The RTM version has been out for over a week. It's only fair especially since you're comparing Ubuntu 12.10 which is very far from being released officially. It's pretty easy to get a Windows 8 ISO from TechNet (I can provide that for you), then get a CDKey from a torrent just to install it. As long as you don't activate the OS, the pirated key shouldn't be a problem. You can go 30 days before you need to activate a Windows 8 installation with a CDKey. Unfortunately, you need to put in a BS key before the install starts unlike Windows 7.
2 versions of OS X, 2 versions of Ubuntu, 2 versions of Windows. Makes perfect sense.
Last edited by elg2001; 08-24-2012 at 12:59 PM. Reason: added last paragraph
I'm really REALLY getting tired of you yapping about your worthless faggotbook.
ENOUGH ALREADY, NOBODY GIVES A SHIT ABOUT GARBAGE!!!!