Quadro 140M troubles

Starting to have a bunch of troubles with my Nvidia Quadro 140M card on a Dell D830 laptop. Main feature is that the screen will either randomly go blank, or will go into severe oscillations in terms of the picture, on both the internal, as well as the external attached LCD.

Is this a symptom of the notorious and well-known defects in the Nvidia mobile graphics chipsets? My laptop is under warranty, so I can get a new motherboard from Dell just for asking -- but I want to be sure this is a hardware thing before I pursue much further.

Anyone?

 

If it's blacking out randomly (or appearing to anyway) check your Power settings and make sure you monitor isn't set to be powered down by the system to save power or for certain events. You should also check your Event Logs to see what's being logged when it goes off.

1. Control Panel, Performance Information
2. Advanced Tools - View Performance Details in Event Log

Check the Admin log in the Custom View folder, the Application, Security and System logs in the Windows Logs folder, and Hardware in the App/Services folder.

If you go back to Advanced Tools and click on the Reliability and Performance Monitor, then look at the Reliability Monitor graphs, it will show you if you've had a hardware failure (they're noted by a red circle with white X)

Also in the AT window at the bottom is a System Health Check diagnostic tool. It will collect data for about a minute, the send the Report to the R&P logs. Look under the Reports/System Diagnostic folders - the report from the scan will be there, the name being the date of the scan. It runs several specific tests on the GPU.

You might also want to install and run HWMonitor if you don't have it (CPUID.com) to see if you've got an overheating problem that's causing your GPU to fail. The symptoms you describe are consistent with GPU overheating, especially the picture oscillations you mentioned.

 

Okay, GPU = 73 degC / 163 degF, just in Windows, nothing running.. Intel T7500 CPU is at 62degC/142degF.

Dual WUXGA screens, so I'd assume that this would naturally lead to higher temperatures than, for instance, with some lower-resolution screens.

 

73ºC while idling is a bit hot. You could try measuring the temperatures while you tax the GPU, like when playing a game.

 

Fixed this problem by blowing the computer out with lots of compressed air.

GPU idle temperature is back in the 40-50 degC range. And crashes and slowness have dissappeared.