First of all: I'm sorry if I'm mistaken about the section to post this question and I'll gladly fix it in case I'm wrong. Now to the issue itself:
I've been testing, for some time, a few Linux distributions (including Ubuntu 16.04). One of those test profiles (qgears2 - Text) yields results way below average for Ubuntu when compared to other distributions. I'd like to know whether those results are natural or, in some way, configuration dependent. What follows is the system under test configuration (all other test were executed in the same machine):
Phoronix Test Suite 6.6.1
Processor: Intel Pentium E6700 @ 3.20GHz (2 Cores)
Motherboard: DIGIBOARD G41M-S v5.0
Chipset: Intel 4 DRAM + ICH7
Memory: 4096MB
Disk: 320GB Seagate ST3320418AS
Graphics: Intel 4 IGP
Audio: VIA VT1708S
Monitor: SMB1630N
Network: Qualcomm Atheros AR8132 Fast + Ralink RT3060 Wireless 802.11n 1T/1R
OS: Ubuntu 16.04
Kernel: 4.4.0-15-generic (x86_64)
Desktop: Unity 7.4.0
Display Server: X Server 1.18.1
Display Driver: intel 2.99.917
Compiler: GCC 5.3.1 20160320
File-System: ext4
Screen Resolution: 1366x768
Unit is frames per second
(Full size available at https://postimg.org/image/o609e1f9n/)
Announcement
Collapse
No announcement yet.
Ubuntu 16.04 Intel Graphics: Unity, Xfce, KDE, LXDE, GNOME, MATE, Openbox
Collapse
X
-
I've read this whole thread and can't see an answer to this: Can someone explain HOW Unity beats Openbox? Given that Openbox won't even be using a compositor?
Leave a comment:
-
Taking into consideration how many dislike unity, it is funny to see it as the overall winner.
Leave a comment:
-
Originally posted by darkbasic View PostYou are confusing the "Suspend compositor for full screen windows" option with manually temporarily disabling compositor with ctrl+shift+F12. The latter is the suggested way by kwin's developer and the one you should use when playing games (of course, it works with Intel).
Originally posted by darkbasic View PostEdit: I personally think that the real solution would be implementing _NET_WM_BYPASS_COMPOSITOR but Martin doesn't care so I fear we will never see it.
Of course, we could get both features, but since developers won't implement both, I'd rather have the manual disabling of the compositor.
Leave a comment:
-
Originally posted by Michael View Post
Each of the desktops was tested out-of-the-box. Plus with Intel anyways, doesn't that option on KDE not have any effect? They had disabled it in the code for Intel to have no effect, or did they end up reverting that decision?
As far I read and remember, automatic undirected composition for full screen window is not working and disabled. But Ctrl+Shift+F12 disable composition manually. It squeeze out some fps. In Dota2 or Team Fortress it is no factor for me(already vblank 60fps) but I have seen improvement in Shadow of Mordor or Metro redux with it.
Leave a comment:
-
Originally posted by Michael View Post
Each of the desktops was tested out-of-the-box. Plus with Intel anyways, doesn't that option on KDE not have any effect? They had disabled it in the code for Intel to have no effect, or did they end up reverting that decision?
Edit: I personally think that the real solution would be implementing _NET_WM_BYPASS_COMPOSITOR but Martin doesn't care so I fear we will never see it.Last edited by darkbasic; 18 April 2016, 04:42 AM.
Leave a comment:
-
It's too bad you didn't have a chance to test the SteamOS compositor in Ubuntu from the https://launchpad.net/~mdeslaur/+archive/ubuntu/steamos PPA,
Would it be tough to get something like this going with your Phoronix test suite. Ya know, with SteamOS being a big picture mode interface.Last edited by tegs; 17 April 2016, 02:57 PM.
Leave a comment:
-
Guest repliedOriginally posted by rabcor View PostWould like to see some Solus benchmarks, they boast pretty loudly about their desktop performance optimizations, would be nice to see if they live up to it.
Leave a comment:
-
So knowing that pb of x11, and that you tested intel graphics... Can't you do another round with unity8 and gnome+wayland?
Leave a comment:
Leave a comment: