PDA

View Full Version : Any difference between OS of same kernel?



inashdeen
June 2nd, 2012, 02:29 PM
Hi there, I would like to ask a question here. is there any difference in terms of kernel performance between two OS, say archlinux and ubuntu when both are using the same version of Kernel? thanks in advance

0011235813
June 2nd, 2012, 02:33 PM
Yes, but not much.
Arch Linux would be slightly faster because it is a lighter base (which is what type of packages a distribution uses and is less important than the kernel) and therefore includes less services, programs running etc.
Especially if you're using LXDE with Arch as opposed to Unity in Ubuntu.

inashdeen
June 2nd, 2012, 02:34 PM
I see a great difference in temperature of the PC when using Arch fork, Chakra in oppose to Ubuntu 12.04 . both are using a similar Kernel. what might be the cause?

0011235813
June 2nd, 2012, 03:49 PM
Might be Chakra is newer and more optimised for your hardware. What hardware are you running it on?

PS: How did you measure it? Software is known to give inaccurate results in some cases.

inashdeen
June 2nd, 2012, 03:51 PM
I am using HP compaq laptop CQ42 203 AU.

random specification, RAM 3GB, ATI graphic card with AMD dual core.
well, there is only 1 thing i check on this laptop since it has been suffering from overheating since 11.04. i installed lm-sensors in both and perform the similar taks in bots OS. I can get up to 97^ on Ubuntu 12.04, but only 90* max on chakra. on average, ubuntu runs at 78* and chakra at 72*

0011235813
June 2nd, 2012, 08:45 PM
I am using HP compaq laptop CQ42 203 AU.

random specification, RAM 3GB, ATI graphic card with AMD dual core.
well, there is only 1 thing i check on this laptop since it has been suffering from overheating since 11.04. i installed lm-sensors in both and perform the similar taks in bots OS. I can get up to 97^ on Ubuntu 12.04, but only 90* max on chakra. on average, ubuntu runs at 78* and chakra at 72*

lm-sensors? That's a software based solution, and I've heard software is notorious for giving inaccurate results in this type of thing. You might get a more accurate reading from something like the phoronix-test-suite or the AMD/ATI control panel. Of course, the best way is to put a thermometer on the thing itself and see what you get.

Anyhow, those differences are quite small. There's no absolute factor I can pinpoint, just that some distributions run better on certain hardware than others.

inashdeen
June 2nd, 2012, 11:16 PM
Hurm,.. you got a point there. how do you actually get the phoronix-test-suite

0011235813
June 3rd, 2012, 10:55 AM
Hurm,.. you got a point there. how do you actually get the phoronix-test-suite

In Debian-based systems like Ubuntu?

sudo apt-get install phoronix-test-suite
Or just search for phoronix in the software center.

Note: The suite you install will ask you to download additional packages for some tests.

inashdeen
June 3rd, 2012, 03:40 PM
so the source is preinstalled. ok, thank you.will try in the near future :) thanks.