6-Way Desktop Comparison On Linux Mint 17

[korrode]

Either combining the Compton compositor (in opengl mode) with Xfwm4 or simply replacing Xfwm4 with your choice of the Compiz 0.8 or 0.9 branch works nicely.

[Viper_Scull]

You can't measure the responsiveness and smoothness with this tests. I really don't care about fps on Xonotic and the likes.
Just playing with them you can see the big difference between heavy DEs and lightweight DEs when it comes to interact with it.
I personally don't use any DE, just Openbox,

[oleid]

Xfce > All

True, except in the cases in which all are basically equal or the GtkPerf test or the first two QGears tests (in which it is as fast as LXDE). Which leaves none of the tests

[philipmorris]

So an i7 u from a ultrabook has the same performance that an Amd Ahtlon 64 of desktop?
Kde has to many bugs with the effects disabled.
What about unity? You can install unity in Mint.
XFCE/KDE has the Windows 95 style so boring for me and old... i have tired to use that on W95,W95 plus, W98, W98se, W2000, WMe, WXP,W7... The gnome shell inferface is nice is keeping better. And better to use gnome because you always load GTK2/GTK3 libraries on the cache. And there are no good aplications on xfce you always have to use gnome apps, like evince filleroller... so you still load gtk3 libraries, if you use KDE you sure use Firefox eclipse... and some components created by de gnome foundations Glib, pango, ...

[amp3030]

And Unity?

Unity does not exist?!

[schmidtbag]

Interesting how KDE was either on top or on the bottom, but never in between. I'm hoping KF5 will help push it away from being the bottom.

[Azrael5]

Micheal could you make the average of all results?

[Vim_User]

XFCE does not have an OpenGL page flipping compositor,

Was added to xfwm4 in the Git already, so will come in XFCE 4.12 or when you compile it yourself from Git.

[halo9en]

If the default is bad, is it not reasonable to change the default, rather than expect every user to know about and to manually change the setting in their personal configuration?

Of course, but every single KDE benchmark I've seen on Phoronix gives users the impression that it's a bloated, windowzian beast that cannot compete with other DE's. I avoided it like the plague and tried GS, various WM's, and finally Cinnamon, until I decided to try KDE again a few days ago. Tweaked a couple of easily accessible settings, run powertop and a few simple benchmarks and frankly, it's a great DE.

And here's a few interesting user benchmarks:

http://www.phoronix.com/forums/showthread.php?83954-Ubuntu-13-10-Desktop-Tests-Unity-7-1-KDE-4-11-Xfce-4-10-GNOME-3-8&p=354515#post354515

http://www.serkey.com/ubuntu-gtkperf-gnome-vs-kde-compiz-vs-kwin-bfkcp5.html

[Artemis3]

Memory usage

What these tests are missing is memory usage, which is one of the major points of using something like LXDE in the first place.
Boot to desktop times would also be interesting.