Windows 7: The nightmare called : "Display driver XYZ stopped responding"

E5200 @ 2.5GHz
Gigabyte GA-73PVM-S2H
Corsair 2 x 2GB DDR2 800MHz
600W PSU which supplies 36A combined on 2 x 12V rails
GeForce GTS 250 1GB (DK version)
Win 7 64bit Ultimate
Bios settings – default, nothing OCed.
When Im in windows browsing internet or listening to the music everything is fine. CPU idles at 36 – 38 deg C, graphics at 40 deg C. Everything is stable and looking really good. I was really happy, liked win 7 as well as my new “old-spec” computer.
I installed COD4 on this machine and the nightmare started. No matter what settings I set in options menu, the game was freezing after 5-10 secs. I could hear the sound still working, but the picture was still. Alt-tab didn’t work. Ctrl+Alt+Del did work, managed to have a look at task manager, switched back to COD4 and it was ok but only for the next 5-10 secs. Then it froze again. I could do this Ctr+Alt+Del thing as many times as I wanted to be able to go back into the game and then “play” it for 5-10 secs. Within this time gap I could go back to menu/change settings/exit game without any problems. But if I messed about with this out-to-windows and back-into-cod4 stuff, after a few “cycles” I got the message from win 7 that the “…Displaydriver XYZ stopped responding and has recovered...” where XYZ is the number of my graphics card’s driver installed. So I uninstalled the drivers for my gfx (in safe mode using sweeper) and installed different ones on. Tried older version of drivers, newer ones, win vista 32/64bit, win 7 32/64bit. It didn’t have any impact on the performance in COD4. Still the same problem. Then I started looking at what happens to my gfx during these “crashes”. GPU load 0% (but read on internet that GPU-Z doesn’t always show it correctly), temp 48 deg C, CPU load 100% on 1 core, 13% on the second one, temps on cpu 53-58 deg C. Tested my graphics card with video card stability testing software – passed (max temp during test 63 deg C), upped the fan speed to 80% just like it was suggested somewhere on web, same problem with COD4. I even did prime95 torture test – stopped it after 10 hours – no problems whatsoever. Then I took my gfx out and tried running COD4 on the onboard gfx – same thing. I installed the COD on my wife’s machine (a bit stronger spec only weaker gfx same Win 7) – no problems whatsoever ! Didn’t try any other games since I haven’t got any more.
Any suggestions please ? Has anyone experienced similar problem and managed to solve it ? Im really frustrated with this now, browsed the internet a lot to find the solution. Some people managed to get this sorted by RMAing their mobo or gfx or just upping the fan speed, manually setting ram timings or reinstalling drivers (for some older drivers worked for others the newish ones – there is no logic behind it) or even going back to vista/xp. Perhaps, the only and in this case the easiest solution, is to swap the machines with my wife…but I would rather get this problem sorted instead of getting simply beaten by it. Thanks.

This error is usually related to bad RAM/GPU
Test your RAM: RAM - Test with Memtest86+
Or consider RMAing the card

The problem persists even with the external gfx card not being inside the computer.
Tested RAM as well as mobo and cpu with prime95 torture test, no issues found.
Will run the Memtest86+ as well as soon as I get back home, but what if it won't display any errors...

Welcome to the forum inhibitor.

These TDR errors can be caused by an array of issues, mostly hardware related. It requires detective work to find the cause.

The nVidia 200 series graphics cards were prone to these issues. While nVidia makes the graphics engine it is the manufacturer (EVGA, PNY, etc) that makes the card. A lot of cards were replaced in what amounted to a silent recall. So don't rule that out.

Because your problem persists whether the on-board or add-in graphics is enabled I think you may be looking at other causes, like RAM, Power Supply, or overheating.

I helped a guy in this forum a while back who ultimately solved his TDR problem by replacing the RAM. Thing is, his old RAM passed Memtest without errors. So while running these tests is an important diagnostic, it is not an end-all.

Take a look at this thread (post #9) and see if you can self diagnose the cause of your issue:
Random freezing/crashing with Nvidia GeForce GT 220

Thanks for the advice !
When I got back home yesterday I decided to have a go with a different mobo (had a spare one knocking about). It was ok, but COD4 was still freezing. Then I installed Win XP Pro SP1 just to check if the problem would still persist with the old OS, but it was the same. However, the second mobo (AMIBIOS) started giving me 2 short beeps on boot, which as I found out later meant that there was a problem with the onboard memory.
So I did run memtest86+ last night, it was giving me plenty of errors with both mem sticks in, then eventually it was crashing. Tested sticks one by one after that, did 13 cycles on one with no errors whatsoever, second one on its own gave me about 146 errors just after the test had started. So I guess, I found the first but not necessary the last faulty part in my machine ! Fingers crossed swapping the ram stick will end up all my problems...

This confirms the RAM problem with one stick, replace it

+1 on that.
It is refreshing to see one of these TDR posts where the problem has a definitive diagnoses.

It is important to replace the bad RAM stick with the exact same model #. If you can't do that, get a matched pair.

Thanks ! Upped your rep ofc !
Btw theoretically atm with only 1 stick (the "good" one) in my mobo everything should work fine...
Will have a go and let you know.

Sorry for dblposting. Just wanted to let you know, that it worked brilliant with the good memory stick. Im returning the bad one to the shop today to get a replacement. Thanks a lot guys one more time. The memtest86+ will stay in my memory for a long time