ATI performance is worse in Linux than in Windows

Hello,
I am using the catalyst driver (version 10.9), with xorg 1.8. Playing some demanding games like 0 A.D shows a lot worse performance on Arch than on Windows (with the same game). My ATI card is an ATI HD 4350.
Is this difference in performance normal? Is there any way to make Arch's graphic processing match Windows's?
I attach my xorg.conf:
Section "ServerLayout"
Identifier "aticonfig Layout"
Screen 0 "aticonfig-Screen[0]-0" 0 0
EndSection
Section "Module"
EndSection
Section "Monitor"
Identifier "aticonfig-Monitor[0]-0"
Option "VendorName" "ATI Proprietary Driver"
Option "ModelName" "Generic Autodetecting Monitor"
Option "DPMS" "true"
EndSection
Section "Device"
Identifier "aticonfig-Device[0]-0"
Driver "fglrx"
BusID "PCI:1:0:0"
EndSection
Section "Screen"
Identifier "aticonfig-Screen[0]-0"
Device "aticonfig-Device[0]-0"
Monitor "aticonfig-Monitor[0]-0"
DefaultDepth 24
SubSection "Display"
Viewport 0 0
Depth 24
EndSubSection
EndSection

Darksoul71 wrote:
@Davigetto:
Hm, sound like fglrx is loaded correctly. Strange. Check dmesg output and your logs if you see any hint on problems.
flamelab wrote:Fglrx is a piece of garbage. You are lucky that you can have a display on Linux with that.
True, but unless radeon gets a big feature and speed boost fglrx is the only option under Linux to get maximum performance.
At least unless you plug a NVidia card in your mobo.
For now even the "bleeding edge" radeon drivers for R600 chips is mostly only usable for older games (Quake, ioquake, UT2k4).
Well, I don't see anything strange in dmesg. In fact, when I said "performance", I meant "30 fps on Windows, 17-20 fps on Linux".
Maybe is a merely driver problem, but I just want to know wheter I am alone in this kind of problem and more people experiences the same.
Last edited by Davigetto (2010-10-03 21:52:58)

Similar Messages

Is the compilation done differently on Linux than on Windows?

Hey guys,
I am currently having some issues with a FPGA program with Softmotion not compiling.
We are running out of options in terms of how to get this FPGA program compiled.
So my question is does compiling on Linux have a different probability of compiling FPGAs?
I have heard that Xilinx Compiler is meant for Linux so it runs more efficiently and faster,
so I was just wondering if the compile method was different as well
Jeff

It doesn't matter if you run it on Windows or Linux. The compiler was just optimized to run on Linux, therefore it can run faster on Linux. But the compilation itself won't matter.
I'd say it is time for you to go through your local NI representative to try to get some time with the NI Systems Engineering group. They will likely find plently of places to help you optimize and fit your code in into an FPGA.
There are only two ways to tell somebody thanks: Kudos and Marked Solutions
Unofficial Forum Rules and Guidelines

Java program consumes far more ram on linux than on windows - why ?

Hello dudes,
i encountered a problem and i dont know how to solve it. i've a java application that consists of several classes and 3 threads. it consumes about 7 mb on win2k but up tp 100 mb on SuSE Linux and shows up 13 threads.
is this something wrong with java or with linux? have any of you guys already faced the same problem and solved it ?
thanks for any replies.

Be careful. The Linux ps command by default shows the individual threads of a Java program as separate processes when they are in fact not. The fact you have 13 threads is normal, most are for housekeeping and appear on Windows as well.
I can't speak to the total memory utilization, but make sure you aren't adding up the memory of each of the 13 lightweight processes (threads) and counting the same memory usage 13 times.
Chuck

Printing from Acrobat Reader Linux is many time slower than from Windows Reader

The users are complaining to me that same file is printing one page every couple minute under Linux, while under Windows pages from the same file print one after another very quickly.
Hardware is not affecting that as they are using dual-boot Fedora 14 and XP on the same machines, printing to the same HP LaserJet printer over JetDirect.

Also, I am noticing that scrolling page thumbnails and flipping pages is many times slower under Linux than under Windows on the same machine, but CPU utilization is very low.
Can one of the Linux Acrobat developers look into performance issues please?

Worse performance with Radeon HD 6750m than Intel 3000?

Lately when i play games (specifically World of Warcraft and Skyrim) and watch movie i get worse FPS and performance on my dedicated GPU than intel 3000. i called apple care and they had me reset the VRAM(?) and something else, i forget. to get rid of the problem i need to restart my computer, but it will resurface after a couple hours.

Lately when i play games (specifically World of Warcraft and Skyrim) and watch movie i get worse FPS and performance on my dedicated GPU than intel 3000. i called apple care and they had me reset the VRAM(?) and something else, i forget. to get rid of the problem i need to restart my computer, but it will resurface after a couple hours.

ATI Catalyst 8.5 Proprietary Linux Display Driver released

New ATI Catalyst™ 8.5 Proprietary Linux Display Driver was released the day before yesterday and according to release notes it has many improvements.
New Features
This release of the Catalyst™ Linux software driver introduces Mipmap Detail, Catalyst A.I., and AA Filters in the 3D ORCA page of the Catalyst™ Control Center Linux Edition.
Resolved Issues
* The following section provide a brief description of resolved issues with the latest version of the ATI Catalyst™ Linux software suite. These include:
* Support for video playback on the second head in dual head mode is now available. Further details can be found in topic number 737-26985
* Unresolved symbols in Xorg log under RHEL 4u6 32 and 64 bit no longer occur. Further details can be found in topic number 737-34284
* The desktop background color is no longer displayed incorrectly when using xcommgr. Further details can be found in topic number 737-34285
* Change VT no longer fails when kernel module is not loaded. Further details can be found in topic number 737-34286
* Switching to the virtual terminal multiple times no longer results in X displaying corruption. Further details can be found in topic number 737-34287
* Catalyst Control Center: The Preference page is now localized for all supported languages. Further details can be found in topic number 737-34275
* The Linux kernel module is no longer installed to wrong location if the Linux kernel is updated. Further details can be found in topic number 737-34288
* Random system hangs no longer occur when calls to the ATI driver are made from Ring3. Further details can be found in topic number 737-34289
* X window no longer fails to respond when GoogleEarth is maximized. Further details can be found in topic number 737-34290
* A segmentation fault no longer occurs when running SPECViewperf on systems containing an ATI FireGL V5100 series product and running SUSE 10.3 x86. Further details can be found in topic number 737-34276
* Enabling Composite extension no longer displays 8x AA for graphic cards that do not support 8x AA under the Ubuntu 7.10 operating system. Further details can be found in topic number 737-34291
* Display corruption is no longer noticed at logon when compiz is enabled. Further details can be found in topic number 737-34292
* Launching the glxgears application no longer results in segmentation faults. Further details can be found in topic number 737-34293
* GoogleEarth: Launching the application in horizontal mode on a system running X no longer results in corruption being noticed. Further details can be found in topic number 737-34293
* Specviewperf 8.1 64 bit version: Segmentation faults no longer occurs on IGP systems. Further details can be found in topic number 737-34294
* Maya 2008: Moving the render window no longer results in the operating system failing. Further details can be found in topic number 737-34295
* Glxgears corruption is no longer noticed in big desktop mode. Further details can be found in topic number 737-34296
Known Issues
* The following section provides a brief description of known issues associated with the latest version of ATI Catalyst™ Linux software suite. These issues include:
* Desktop corruption may be noticed when dragging the overlay/video when using dual-display mode. Further details can be found in topic number 737-29578
* Bandwidth corruption may be noticed when setting the desktop resolution to 2048x1536 on systems containing an ATI Radeon HD 3200 series of product. Further details can be found in topic number 737-33499
* SECAM, NTSC or PAL signal changes when using aticonfig or Catalyst Control Center but does not apply to the TV display output. Further details can be found in topic number 737-34297
* Running X server without a window manager may result in the X server flashing green for a moment, going black, and the computer becoming unresponsive. Further details can be found in topic number 737-34299
* Desktop corruption may be noticed when dragging the overlay/video when using dual-display mode. Further details can be found in topic number 737-34298
* Connecting a CRT display device to a system containing an ATI Mobility Radeon X1600 product may result in the display device flashing multiple times or failing to display an image when attempting to launch the Catalyst Control Center. Further details can be found in topic number 737-34300
* Corruption may be noticed when connecting a DFP display device to a system running X after setting the TV mode to 1080i. Further details can be found in topic number 737-34301
* Quake 3 Arena (demo): Segmentation faults may occur when attempting to play the game. Further details can be found in topic number 737-34302
* Specviewperf: Launching the application may result in the Linux operating system failing to respond. Further details can be found in topic number 737-34282
* Opening a terminal and running semi-automated DPPT tests via lnxtest.sh may result in the operating system failing to respond. Further details can be found in topic number 737-34283
* A segmentation fault may occur when running SPECViewperf on systems containing an ATI FireGL V5100 series product and running SUSE 10.3 x86. Further details can be found in topic number 737-34319
* FGL_GLXGEARS or GLXGEARS: X may fail to respond when launching either of the applications. Further details can be found in topic number 737-34316
* Maya 2008: Various comparison failures may be noticed. Further details can be found in topic number 737-34277
* The X Server may fail to reset clearly when AIGLX is enabled. Further details can be found in topic number 737-34317
* DiscoveryStudio1.7: The print preview and save as functions may fail to work properly. Further details can be found in topic number 737-34278
* SPEC Viewperf 8.1: A performance drop may be noticed when running the application under Linux. Further details can be found in topic number 737-34279
* Maya2008: The hardware render buffer may fail to render/display the image. Further details can be found in topic number 737-34280
* May2008: The hardware render Perl script may display an error message when rendering. Further details can be found in topic number 737-34281
I wonder when it will appear in Arch's repositories

Yea, it would be great as at Phoronix forums some people are reporting 10% performance increase in games
EDIT: Some release notes from AMD
New Features
This release of the Catalyst™ Linux software driver introduces UYVY and YUY2 pixel format support. This support provides interleaved stream support for video playback applications such as TVTime and MythTV.
Resolved Issues
The following section provide a brief description of resolved issues with the latest version of the ATI Catalyst™ Linux software suite. These include:
*Enemy Territory: Quake Wars: Playing the demo on a system running X no longer results in corruption being noticed. Further details can be found in topic number 737-35208
*Maya 2008: An incorrect colormap is no longer displayed for texture selector on systems running X. Further details can be found in topic number 737-35207
*Quake 3 Arena (demo): Segmentation faults no longer occur when attempting to play the game. Further details can be found in topic number 737-35206
*Quake3: Corruption is no longer noticed when changing the display resolution when the game is running. Further details can be found in topic number 737-35209
*SECAM, NTSC or PAL signal changes when using aticonfig or Catalyst Control Center and now gets applied to the TV display output. Further details can be found in topic number 737-34297
*Running X server without a window manager no longer results in the X server flashing green for a moment, going black, and the computer becoming unresponsive. Further details can be found in topic number 737-34299
*Corruption is no longer noticed when connecting a DFP display device to a system running X after setting the TV mode to 1080i. Further details can be found in topic number 737-34301
*Clock Gating is now supported and enable under the Linux operating system. Further details can be found in topic number 737-35191
*The desktop background color is no longer displayed incorrectly when using xcommgr. Further details can be found in topic number 737-35192
*Using the wrong delimiter in aticonfig --set-dispattrib=<disp>,<attr>, <value> no longer results in a segmentation fault. Further details can be found in topic number 737-35194
*Audio out is now functioning after the installation of the ATI graphics drivers for Ubuntu 7.10. Further details can be found in topic number 737-35196
*Bandwidth corruption is no longer noticed when setting the desktop resolution to 2048x1536 on systems containing an ATI Radeon HD 3200 series of product. Further details can be found in topic number 737-33499
*Playing a video clip using mplayer with the Xv extension enabled no longer results in an error indicating that Xv could not find a free Xv port. Further details can be found in topic number 737-35198
*Corruption is no longer noticed when enabling Big Desktop on systems containing an ATI FireGL X1 series of product and running X. Further details can be found in topic number 737-35199
*Running the X server generations test no longer results in the operating system failing to respond or a Kernel panic to occur after first server generation on SuSE 10.3 64 bit. Further details can be found in topic number 737-35200
*Attempting to resize a 2D application window no longer results in poor performance being noticed when composing is enabled. Further details can be found in topic number 737-35201
*Screen flicker is no longer noticed after running amdcccle in X. Further details can be found in topic number 737-35202
*Corruption is no longer noticed when connecting a DFP display device to a system running X after setting the TV mode to 1080i. Further details can be found in topic number 737-35203
*Glxgears corruption is no longer noticed in big desktop mode. Further details can be found in topic number 737-35204
*Starting glxgears and/or fgl_glxgears no longer results in the gears failing to render on SuSE 10.2 x86_64. Further details can be found in topic number 737-35205
*FGL_GLXGEARS or GLXGEARS: X no longer fails to respond when launching either of the applications. Further details can be found in topic number 737-35210
Known Issues
The following section provides a brief description of known issues associated with the latest version of ATI Catalyst™ Linux software suite. These issues include:
*DiscoveryStudio1.7: The print preview and save as functions may fail to work properly. Further details can be found in topic number 737-35219
*Desktop corruption may be noticed when dragging the overlay/video when using dual-display mode. Further details can be found in topic number 737-29578
*Bandwidth corruption may be noticed when setting the desktop resolution to 2048x1536 on systems containing an ATI *Radeon HD 3200 series of product. Further details can be found in topic number 737-33499
*Connecting a CRT display device to a system containing an ATI Mobility Radeon X1600 product may result in the display device flashing multiple times or failing to display an image when attempting to launch the Catalyst Control Center. Further details can be found in topic number 737-34300
*DiscoveryStudio1.7: The print preview and save as functions may fail to work properly. Further details can be found in topic number 737-34278
*Setting the Overscan to an invalid value may result in segmentation faults. Further details can be found in topic number 737-35211
*An incorrect error message may be returned when applying an unsupported TV geometry in a console terminal. Further details can be found in topic number 737-35212
*Desktop corruption may be noticed when dragging the overlay/video when using dual-display mode. Further details can be found in topic number 737-35213
*An X segmentation fault may occur when launching SUZI/textured video. Further details can be found in topic number 737-35214
*Creating two successive connections may result in direct GL context to fail. Further details can be found in topic number 737-35215
*Moving the mouse or tapping a key may fail to close an OpenGL screen-saver and bring the user back to the x desktop. Further details can be found in topic number 737-35216
*The X Server may fail to reset clearly when AIGLX is enabled. Further details can be found in topic number 737-35217
*Opening a terminal and running semiautomated DPPT tests via lnxtest.sh may result in the operating system failing to respond. Further details can be found in topic number 737-35218
Last edited by Zibi1981 (2008-06-19 15:42:23)

How can i prove rac on linux is better than on windows ?

Hi all,
Is there any document or any thread on this forum which will help me prove that RAC on linux is better than on windows. Actually i did not recommend to my management windows os for RAC, now the management wants the reason why did not recommend windows. I want some sort of proof or any document to prove my point. Please help ......

> now the management wants the reason why did not recommend windows
Then give them reasons. TCO. In-house experience and skills with Linux. Official support from vendors like Sun for Linux on their servers. Linux being used as the core development platform for Oracle. Linux proven as a cluster-based operating system - it leads the supercomputer/cluster market by a significant percentage.
According to www.top500.org (listing the 500 fastest/biggest computers clusters on this planet)*:
Operating System Family Top500
Linux 85.20 %
Mixed 6.80 %
Unix 6.00 %
Windows 1.20 %
BSD Based 0.40 %
Mac OS 0.40 %
* (In other words, if Linux is good enough to run 85% of the world's fastest computers, it is more than good enough to run your company's RAC)
There are numerous sound reasons for choosing Linux over Windows. But do not make performance claims that you cannot ever hope to backup with technical evidence.

Perform VENDOR EVALUATION for MORE THAN ONE VENDORS at a time

Hello all,
Please guide for any process where i can perform Vendor Evaluation for MORE THAN ONE vendors AT A TIME.
At my location there are around thousand vendors, which are to be evaluated, and difficult to perform the evaluation process one-by-one.
(ME61/ME62/ME63)
Detailed replies with various possibilities would be highly appreciated.
Thanks & Regards,
Joy Ghosh

The vendor evaluation for some thousand vendors at the same time has already been in SAP long before they developed LSMW. The purpose of LSMW is to load data from a legacy system, of course you can (mis-)use it for a lot other things.
But you should not always use LSMW if you are to lazy to go thru the SAP standard menu to find a transaction like ME6G
There you define a job that runs RM06LBAT report.
You first have to define a selection variant for this report. this can be done in SE38 by entering the report name, select variant, clicking display, then entering a name for the variant and clicking Create.

ORACLE performance get worse after 24 hours of inactivity

Yesterday I imported a ORACLE DB, and I executed a query which uses some indexes. The query had only a LEFT JOIN between two tables. 24 hours later I executed the same query and then ORACLE decided not using the indexes and now uses full table access. That is the reason because the performance is worse. Does anybody know what is happening?

Did you note when you posted this that the forum name is "Adevanced Queueing?"
Please repost your question in the appropriate forum and include full version information, DML, and an explain plan when it is working properly and another from when it is slow using DBMS_XPLAN to generate the output.
You can find demos of how to do this in Morgan's Library at www.psoug.org under EXPLAIN PLAN.
Without this information no one will be able to help you.
Again ... do not post it here ... post it in the correct forum.

GetTimezoneOffset got one hour ahead on linux than windows

I'm using Date object to retrieve current local time for
display. But got one hour ahead than the system time.
For an example, I set the system time zone to Hong Kong,
China. The getTimezoneOffset() return -540 on linux, but on
Windows, it's -480. So, the time in flash is one hour ahead than
the system time.
I have checked that the getUTCHours() return the same result
on linux and windows.
Any Suggestions?
thx!

Can somebody please confirm/deny this.
FP 9 on Linux shows time one hour ahead than the local time.
Try following line with FP9 on Linux
trace("Current Time "+new Date());

Why performance of Illustrator CS6 slower than Illustrator CS5.5?

Ex: Process get text of TextFrame has 657 characters:
Dim t1 As String
Dim t2 As String
t1 = DateTime.Now.Minute & " " & DateTime.Now.Second & " " & DateTime.Now.Millisecond
Dim text As String
For iChar = 1 To iCharNum
      oChar = oContents.Characters(iChar)
      Dim str As String
      str = oChar.contents.ToString
      text = text & str
Next
t2 = DateTime.Now.Minute & " " & DateTime.Now.Second & " " & DateTime.Now.Millisecond
Result:
Time of CS5.5: 0s 487
                         t1= 38 43 71
                         t2= 38 43 558
Time of CS6: 3s 871
                         t1=42 54 738
                         t2=42 58 609
Why performance of Illustrator CS6 slower than Illustrator CS5.5?
Can i set attribute increase performance of Illustrator CS6 ?

Hi, yes
CS5 - I run it several times, consistently getting 1 second
9/23/2014 22:00:09.000
9/23/2014 22:00:10.000
CS4 - the script finished immediately, super fast, in a fraction of a second
9/23/2014 22:07:20.000
9/23/2014 22:07:20.000

Firefox will not open more than one window and upon restarting will not open the previous tab and refuses to open bookmarks

Firefox does not open more than one window. I can open plenty of tabs, but not more than one window. Upon closing firefox and trying to restart the browser will not open and needs to be closed through windows task manager. Upon trying to re-open Firefox I receive the 'oops this is embarassing , firefox cannot re-open your tabs" message. When I click on "new session" the bookmarks do not load. It is becoming very frustrating as I am a student and often do lots of research on the web while using lots of open windows.

Start Firefox in [[Safe Mode]] to check if one of the add-ons is causing the problem (switch to the DEFAULT theme: Tools > Add-ons > Themes).
* Don't make any changes on the Safe mode start window.
See:
* [[Troubleshooting extensions and themes]]
* [[Troubleshooting plugins]]
If it does work in Safe-mode then disable all extensions and then try to find which is causing it by enabling one at a time until the problem reappears.
* Use "Disable all add-ons" on the [[Safe mode]] start window to disable all extensions.
* Close and restart Firefox after each change via "File > Exit" (Mac: "Firefox > Quit"; Linux: "File > Quit")
See also "Hang at exit":
* http://kb.mozillazine.org/Firefox_hangs
* [[Firefox hangs]]

Unable to minimize windows or view more than one window

I had a MacBook Pro prior to the MacBook 2015. I never had a problem minimizing windows. I really need to have more than one window open at same time, but I can't seem to do this with MacBook. There usually is a white dot in middle where it was yellow in MacBook Pro to minimize the window. There is an icon with double page symbol I can click in the right corner, but when it shows a thumbnail image of each website on one page instead of minimizing the window when I click on it. Is there something I can click on to minimize the window so more than one window appears at the same time? I do not even see line/arrow appear in corners of window to drag the window to a smaller size.

I do not even see line/arrow appear in corners of window to drag the window to a smaller size.
Just put your cursor on any edge or corner of a window until you see the double arrows then click and hold while you drag to resize. This is no different than it's been for the last 4 years since the introduction of Lion.
There usually is a white dot in middle where it was yellow in MacBook Pro to minimize the window.
The yellow dot will minimize the window to the Dock.
http://www.peachpit.com/articles/article.aspx?p=412362&seqNum=6
In Yosemite, a click on the the Green dot goes to full screen, an Option click is like the prior green dot's performance.

With more than one window open in firefox 4 I starts to drain my memory

With mozilla 4 I have more than one window open and my memory which has 2 gigs in it starts to drain and then everything freezes.

Which Windows platform shouldn't make much difference. The screenshot may look as coming from an older Windows version, but the basics should be the same.
If it does work in Safe-mode then disable all extensions and then try to find which is causing it by enabling one at a time until the problem reappears.
*Use "Disable all add-ons" on the [[Safe mode]] start window to disable all extensions.
* Close and restart Firefox after each change via "File > Exit" (Mac: "Firefox > Quit"; Linux: "File > Quit")
*https://support.mozilla.com/kb/Troubleshooting+extensions+and+themes

Just upgraded to 9.4, safari will not let me keep more than 3 windows open, if i try to open 4th, one goes away

Safari will not let me keep more than 3 windows open, when I try to open 4th, one goes away, just installed 9.4 last week, nothing but problems

Please read this whole message before doing anything.
This procedure is a test, not a solution. Don’t be disappointed when you find that nothing has changed after you complete it.
Step 1
The purpose of this step is to determine whether the problem is localized to your user account.
Enable guest logins* and log in as Guest. Don't use the Safari-only “Guest User” login created by “Find My Mac.”
While logged in as Guest, you won’t have access to any of your documents or settings. Applications will behave as if you were running them for the first time. Don’t be alarmed by this behavior; it’s normal. If you need any passwords or other personal data in order to complete the test, memorize, print, or write them down before you begin.
Test while logged in as Guest. Same problem?
After testing, log out of the guest account and, in your own account, disable it if you wish. Any files you created in the guest account will be deleted automatically when you log out of it.
*Note: If you’ve activated “Find My Mac” or FileVault, then you can’t enable the Guest account. The “Guest User” login created by “Find My Mac” is not the same. Create a new account in which to test, and delete it, including its home folder, after testing.
Step 2
The purpose of this step is to determine whether the problem is caused by third-party system modifications that load automatically at startup or login, by a peripheral device, by a font conflict, or by corruption of the file system or of certain system caches.
Please take this step regardless of the results of Step 1.
Disconnect all wired peripherals except those needed for the test, and remove all aftermarket expansion cards, if applicable. Start up in safe mode and log in to the account with the problem. You must hold down the shift key twice: once when you turn on the computer, and again when you log in.
Note: If FileVault is enabled, or if a firmware password is set, or if the startup volume is a software RAID, you can’t do this. Ask for further instructions.
Safe mode is much slower to start up and run than normal, with limited graphics performance, and some things won’t work at all, including sound output and Wi-Fi on certain models. The next normal startup may also be somewhat slow.
The login screen appears even if you usually log in automatically. You must know your login password in order to log in. If you’ve forgotten the password, you will need to reset it before you begin.
Test while in safe mode. Same problem?
After testing, restart as usual (not in safe mode) and verify that you still have the problem. Post the results of Steps 1 and 2.

ATI performance is worse in Linux than in Windows

Similar Messages

Maybe you are looking for