11.2milestone 6 vrs Ubuntu 9.10Alpha benchmarks

http://www.phoronix.com/scan.php?pag...x_aug_09&num=1
--
openSuse 11.1 x64bit, KDE4.x Factory, Opera weekly

[Dexter1979]

Good to see things will run faster on a 'normal' pc since they're using an Intel Atom for the benchmarks.

[oldcpu]

and the conclusion ...

Depending upon the test scenario, different leaders came out on top. For the most part, all three distributions performed roughly the same, which is not that surprising since all of the core packages are the same between Ubuntu 9.10 Alpha 4, OpenSuSE 11.2 Milestone 6, and Mandriva Linux 2010.0 Beta 1.

While its neat to see such a comparision, I also have an opinion that speed differences between major distributions is not the criteria against which they should be evaluated when trying to choose a distribution for installation.

If one is looking for speed, then other light weight distributions should be considered instead of Fedora, Mandriva, Ubuntu, or openSUSE.

oldcpu wrote:

>
> and the conclusion ...
>> Depending upon the test scenario, different
leaders came out on top.
>> For the most part, all three distributions
performed roughly the
>> same, which is not that surprising since all
of the core packages are
>> the same between Ubuntu 9.10 Alpha 4, OpenSuSE
11.2 Milestone 6, and
>> Mandriva Linux 2010.0 Beta 1.
> While its neat to see such a comparision, I
also have an opinion that
> speed differences between major distributions
is not the criteria
> against which they should be evaluated when
trying to choose a
> distribution for installation.
>
> If one is looking for speed, then other light
weight distributions
> should be considered instead of Fedora,
Mandriva, Ubuntu, or openSUSE.
>
>
but it is a consideration to others and thus it
is important for a distro to at least generate
comparable test results if it wants to have it's
community grow.


--
openSuse 11.1 x64bit, KDE4.x Factory, Opera
weekly

[Confuseling]

oldcpu; well said.

google01103; the thing with Phoronix is they seem to have absolutely no understanding of scientific method. There's no point doing default installs, and then comparing them - you need to isolate variables. A vanishingly small proportion of users who are actually interested in speed will do default installs. This test is a case in point - it might have meant something if they'd at least used the same filesystem on all three distros.

Occasionally Phoronix do a decent test, but they seem to come across them by luck rather than design, which is a shame, because it could be really useful data.

[Dexter1979]

Originally Posted by oldcpu

While its neat to see such a comparision, I also have an opinion that speed differences between major distributions is not the criteria against which they should be evaluated when trying to choose a distribution for installation.

Very true. Speed is probably more hardware dependent. I am still puzzled why they ran it on a netbook tho! If you want to really test the speed of a distro run it on a super fast processor Coe-I7 system with loads of RAM and a SSD hard drive

Netbooks should be running Moblin or something if you want a speedy netbook

[dragonbite]

I didn't read it all, but overall it looks like they are more or less on par with each other, no clear winner.

While speed isn't going to be a major factor in picking one distribution over the other, it is a consideration for users to include when looking at the whole package of things.

[oldcpu]

Originally Posted by dragonbite

While speed isn't going to be a major factor in picking one distribution over the other, it is a consideration for users to include when looking at the whole package of things.

We are talking incredibly minute speed differences. IMHO if speed is any sort of a factor, then one should immediately do away with openSUSE, Fedora, Mandriva, Ubuntu, ... all those "heavy" distributions, and go with a much lighter weight distribution.

But each to their own reason for picking one distribution over another.

[consused]

Originally Posted by Dexter1979

If you want to really test the speed of a distro run it on a super fast processor Coe-I7 system with loads of RAM and a SSD hard drive

Netbooks should be running Moblin or something if you want a speedy netbook

Maybe, but for comparative purposes and to "see" differences in speed, you have a better chance on more modest H/W.

> But each to their own reason for picking one distribution over another.

i always go for the distro which can make the cube spin the fastest

even if it crashes every 5 minutes--eye candy is the MOST useful
feature of any computer.

--
goldie