Sign in with
Sign up | Sign in
Your question

Yet another "Driver has stopped working thread"

Last response: in Graphics & Displays
Share
August 27, 2012 11:19:38 AM

Hello guys n' gals,

I've been through most of your threads here to try and resolve the issue already, but no luck.

I bought a new gaming rig a few days ago and assembled it on Thursday, just in time for that sweet sweet Guild Wars 2 pre-purchase headstart. Needless to say, I was devastated when, despite the high-quality graphics that I was receiving, my video kept crashing and flickering to black, followed by the message "Display driver AMD driver has stopped responding and has recovered."; At first, I got these every 10-20 minutes or so, but now I can't play for 5 minutes without getting one, sometimes a whole chain one straight after the other. I've been searching for a solution for several days now, but I can't seem to find anything that works. I now humbly turn to the internet to see if anyone can suggest something.

The parts I bought:

RAM - Adata XPG Gaming v1.0 8GB (2x4GB) DDR3 PC3-12800C9 1600MHz Dual Channel Kit
PSU - OCZ ZS Series 550W '80 Plus Bronze' Power Supply
MB - ASRock Z77 Pro4-M Intel Z77 (Socket 1155) DDR3 Micro ATX Motherboard
Processor - Intel Core i5-3570K 3.40GHz (Ivybridge) Socket LGA1155 Processor (77W)
Video Card - HIS HD 7850 IceQ X 2048MB GDDR5 PCI-Express Graphics Card

A couple of my own parts, reused:

An old 300GB HDD An old DvD/CD Drive


So far, I have tried the following fixes, nothing works:

Installing motherboard drivers.
Installing/updating AMD drivers.
Complete uninstall (using driver sweeper) of AMD drivers, then a fresh install.
Disabling realtek onboard sound drivers.
Disabling hardware acceleration for firefox/adobe flash.
Disabling antivirus software (MSE).
Increasing TdrDelay in the registry from 2 to 8.
Underclocking my GPU. Overclocking my GPU.
Temperature check (everything running great).
Increasing idle clock settings. Disabling Cataclyst Control Center.
Switching my RAM (tried with only 1 stick, switched RAM to slots 2 and 4 from 1 and 3).
Checking wire connections.
Disabling forced antialiasing.
Disabling AMD overdrive.
Disabling tessellation.
Turning game graphics to low.
Probably some others I've forgotten.


Possible solutions I've not tried yet.

Power supply is too weak? - I've been told this could be a cause, but it's difficult to test this without getting hold of a new PSU.


About the occurrence:

Only happens when gaming (Videos, etc. work fine), usually in graphically demanding areas, but not always. It is not confined to 1 game (I've tried it with both WoW and Guild Wars 2 and it appears in both).

I tried disabling TDR through regedit (TdrLevel set to 0). This resulted in hard crashes.

I have tried playing with the monitor plugged into my motherboard and thus using native graphics, and the problem did not appear. However, native graphics are awful. This makes me think that it is GPU related. (hardware or software?)

That's pretty much it. If anyone has any questions/suggestions/anything I could try, please post them. Thanks!
August 27, 2012 1:23:53 PM

To test your PSU, download OCCT from here http://www.ocbase.com/index.php/download and run the Power Supply test of OCCT while keeping an eye on voltage and temperature readings in OCCT window.

Hope this helps.
August 27, 2012 8:06:48 PM

Thank you for your reply, Suat, but I'm a bit of a noob at all this. What should I be keeping an eye out for? I've noticed the power test doesn't give me a reading on my GPU's performance - all the values seem to be for my CPU (4 cores, from 0 to 3) and some voltage figures. I'm not particularly worried about the temperature (fans can definitely handle anything, the CPU is maxing at about 45-50 degrees even at full load), but I have no idea how to understand the voltage figures.
Related resources
August 28, 2012 11:06:22 AM

For GPU performance, run the GPU test. The important parameter for voltage is +12 V and 3.3 V during the test. The variation on these voltages should be less than +/- 5%.
August 28, 2012 11:28:25 AM

Thank you, suat. I've been running the tests and it looks okay on the 3.3V part, although I'm not seeing a +12V anywhere. Is this the cumulative total or is it specified elsewhere? I've taken a screenshot and was wondering if anyone could make any sense of it for me?



I have noticed some strange spikes in the other voltages (vin 4, vin 0 and CPU vcore), but AVCC and 3VCC stayed pretty much around 3.32-3.39 for the whole duration.
August 28, 2012 2:19:16 PM

"Driver has stopped working" isn't caused by a power supply. If the PSU is bad the computer will either reboot or shut off.

The whole "driver stopped responding and has recovered" is indicative of a bad driver install, bad overclock, defective card and rarely bad RAM.

As far as replacing your power supply? Why? Wherever you got your info that it may be the PSU is flat out wrong. In fact it's a good one.

http://www.jonnyguru.com/modules.php?name=NDReviews&op=...

Quote:
Over all, the OCZ ZS series 550W power supply is a very good power supply, albeit a basic power supply. The actual score of this unit came out to 8.7, which easily gets rounded up to 9 since the reason it didn't get a higher score is because of the "value" grade. One of two things could have made this power supply a sure fire "jonnyGURU Recommended" power supply without any doubts or any rounding up of scores: If the unit was modular at the $79.99 price point or if the hard price was $10 to $20 cheaper than it is. Like I said, there is a $20 MIR on this unit right now, so if you do end up scoring this power supply for a mere $59.99, you've done fine!


Some other reviews:

http://www.kitguru.net/components/power-supplies/zardon...

http://www.technic3d.com/review/netzteile/1259-ocz-zs-5...


August 28, 2012 3:54:54 PM

In OCCT 4.3.1, +12 V is fourth item under CPU VCORE. I am talking about OCCT 4.3.1 window when running on my computer. If you do not see it, I do not know why.
August 28, 2012 8:46:34 PM

RussK1 said:
"Driver has stopped working" isn't caused by a power supply. If the PSU is bad the computer will either reboot or shut off.

The whole "driver stopped responding and has recovered" is indicative of a bad driver install, bad overclock, defective card and rarely bad RAM.


Ah, that's a relief. One less part I have to check out, so I guess it's either defective card (most likely) or bad RAM, although I've already tried reseating the ram (it's in slots 2 and 4 now, rather than 1 and 3) and tried it with 1 stick, etc., still giving me the same result. Also, when I plug the monitor directly into the motherboard and use Intel onboard graphics, the problem doesn't occur so I guess it is unlikely to be RAM related.

So most likely the card. Guess I'll have to RMA that one then.

suat said:
In OCCT 4.3.1, +12 V is fourth item under CPU VCORE. I am talking about OCCT 4.3.1 window when running on my computer. If you do not see it, I do not know why.


Absolutely no idea, sorry.

August 28, 2012 10:45:12 PM

Try this, my own personal way of uninstalling/reinstalling drivers.

1) Download Driver Fusion HERE.
2) Download Catalyst 12.8 drivers HERE. but don't install.
3) Download 12.7 CAP3 HERE.
4) Custom uninstall all AMD software through Control Panel, select everything and reboot.
5) Run Driver Fusion and clean AMD display drivers (and nVidia display drivers if you had an nVidia card prior) - reboot.
6) Re-install CCC and reboot.
7) Install the CAP profile and reboot.
!