Go Back   nV News Forums > Linux Support Forums > NVIDIA Linux

Newegg Daily Deals

Reply
 
Thread Tools
Old 03-20-12, 04:17 PM   #1
tmehrkam
Registered User
 
Join Date: Mar 2012
Posts: 3
Default NVS Quadro 450 poor performance.

I am trying to upgrade from an obsolete NVS Quadro 440 video card.

We are running 64 bit RedHat EL 5.4 on a Dell R710 server. The system is part of a data acquisition system running on research vessels world wide.

The NVS 440 card works fine in our application but the NVS 450 falls flat on it's face when we run one of our apps.

The app displays 12,000 little bar graphs displaying RMS data from 12,000 sensors. This data is updated every 512 ms.

We are using QT 4.6.2 to develop our displays. The problem is when the gain is increased so that the signals clip. The display gets very unresponsive.

Any IDEA why the NVS 450 would show significantly worse performance than the NVS 440. We tried and saw the same thing with the NVS 420.

We are driving 4 displays. Running twin-view and Xinerama. To map 4 monitors into one X window.

------------------------------------ Xorg.conf
# nvidia-xconfig: X configuration file generated by nvidia-xconfig
# nvidia-xconfig: version 256.53 (buildmeister@builder97.nvidia.com) Fri Aug 27 20:55:22 PDT 2010

# nvidia-settings: X configuration file generated by nvidia-settings
# nvidia-settings: version 256.53 (buildmeister@builder97.nvidia.com) Fri Aug 27 20:55:01 PDT 2010
# Xorg configuration created by system-config-display

Section "ServerLayout"


Identifier "Default Layout"
Screen 0 "Screen0"
Screen 1 "Screen1" RightOf "Screen0"
InputDevice "Mouse0" "CorePointer"
InputDevice "Keyboard0" "CoreKeyboard"
Option "Xinerama" "1"
EndSection

Section "InputDevice"

# generated from default
Identifier "Mouse0"
Driver "mouse"
Option "Protocol" "auto"
Option "Device" "/dev/input/mice"
Option "Emulate3Buttons" "no"
Option "ZAxisMapping" "4 5"
EndSection

Section "InputDevice"
Identifier "Keyboard0"
Driver "kbd"
Option "XkbModel" "pc105"
Option "XkbLayout" "us"
EndSection

Section "Monitor"
Identifier "Monitor0"
VendorName "Unknown"
ModelName "Unknown"
HorizSync 28.0 - 33.0
VertRefresh 43.0 - 72.0
Option "DPMS"
EndSection

Section "Monitor"
Identifier "Monitor1"
VendorName "Unknown"
ModelName "Unknown"
HorizSync 28.0 - 33.0
VertRefresh 43.0 - 72.0
Option "DPMS"
EndSection

Section "Device"
Identifier "Device0"
Driver "nvidia"
VendorName "NVIDIA Corporation"
BoardName "Quadro NVS 450"
BusID "PCI:12:0:0"
EndSection

Section "Device"
Identifier "Device1"
Driver "nvidia"
VendorName "NVIDIA Corporation"
BoardName "Quadro NVS 450"
BusID "PCI:13:0:0"
EndSection

Section "Screen"
Identifier "Screen0"
Device "Device0"
Monitor "Monitor0"
DefaultDepth 16
Option "TwinView" "True"
Option "MetaModes" "nvidia-auto-select, nvidia-auto-select"
SubSection "Display"
Depth 16
EndSubSection
EndSection

Section "Screen"
Identifier "Screen1"
Device "Device1"
Monitor "Monitor1"
DefaultDepth 16
Option "TwinView" "True"
Option "MetaModes" "nvidia-auto-select, nvidia-auto-select"
SubSection "Display"
Depth 16
EndSubSection
EndSection

Section "Extensions"
Option "Composite" "Disable"
EndSection

Last edited by tmehrkam; 03-20-12 at 04:25 PM. Reason: Add more information about configuration
tmehrkam is offline   Reply With Quote
Old 03-20-12, 05:44 PM   #2
Plagman
NVIDIA Corporation
 
Plagman's Avatar
 
Join Date: Sep 2007
Posts: 254
Default Re: NVS Quadro 450 poor performance.

Can you please attach an nvidia-bug-report.log as per the sticky instructions relative to reporting bugs?

For a performance issue like that, I probably won't be able help much unless I can reproduce the issue myself. Can your application run standalone, or can the display code be extracted such that the problematic scenario can be made into a standalone testcase?
Plagman is offline   Reply With Quote
Old 03-21-12, 01:04 PM   #3
tmehrkam
Registered User
 
Join Date: Mar 2012
Posts: 3
Default Re: NVS Quadro 450 poor performance.

This is a small part of a huge application. It would take a few days to generate a stand alone test case.

I am not sure I will be allowed to spend the time to generate a test case.

I have been instructed try an ATI card or come up with some type of work around. I will see what I can do in the background. It would be nice to stick with the NVIDIA card so we do not have to switch drivers. Switching drivers will be a support nightmare. These things are on Boats all over the world.

One thing I have noticed is the pixel clock on the 450 card is 160 MHZ where the 440 is 400 MHZ. We are driving DVI monitors. Could that be the issue.

qtgears and qtperf4 show the new card to be slower. gtgears twice as slow. gtperf4 not so much.
tmehrkam is offline   Reply With Quote
Old 03-21-12, 01:19 PM   #4
Plagman
NVIDIA Corporation
 
Plagman's Avatar
 
Join Date: Sep 2007
Posts: 254
Default Re: NVS Quadro 450 poor performance.

The pixel clock shouldn't have an impact on rendering performance. I suspect this is not an issue with raw horsepower of the new card; software fallbacks are very expensive on chips newer than NV4x for various reasons, which affects the 450. Generally we try to accelerate all drawing operations, but there might be some cornercases that still cause us to fall back to software, introducing potential performance pitfalls on these newer cards. For this reason unless I can look at the exact sequence and combination of drawing calls, I can't really be of much help. The qtgears issue you're observing might or might not be related; I'll take a quick look at that to see if there's low hanging fruit there.

If putting together a test case isn't an option given the time constraints, can you provide NVIDIA with access to the full application?
Plagman is offline   Reply With Quote
Old 03-21-12, 02:04 PM   #5
tmehrkam
Registered User
 
Join Date: Mar 2012
Posts: 3
Default Re: NVS Quadro 450 poor performance.

Quote:
Originally Posted by Plagman View Post
The pixel clock shouldn't have an impact on rendering performance. I suspect this is not an issue with raw horsepower of the new card; software fallbacks are very expensive on chips newer than NV4x for various reasons, which affects the 450. Generally we try to accelerate all drawing operations, but there might be some cornercases that still cause us to fall back to software, introducing potential performance pitfalls on these newer cards. For this reason unless I can look at the exact sequence and combination of drawing calls, I can't really be of much help. The qtgears issue you're observing might or might not be related; I'll take a quick look at that to see if there's low hanging fruit there.

If putting together a test case isn't an option given the time constraints, can you provide NVIDIA with access to the full application?
I added an attachment with the bug report.

As far as access to our lab system. that might be arranged. Will have to see how to get you through our corp firewall.

The application is difficult to install. Many dependencies involved. Postgresql etc. Plus a lot of configuration of the application and the hardware simulator.

Would take some doing. I will have to discuss this with MGMT and IS to see what can be done.
Attached Files
File Type: gz nvidia-bug-report.log.gz (97.4 KB, 35 views)
tmehrkam is offline   Reply With Quote
Reply


Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


All times are GMT -5. The time now is 04:07 PM.


Powered by vBulletin® Version 3.7.1
Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Copyright 1998 - 2014, nV News.