Not sure what's wrong but I think it's either my PSU or my GPU

CSDragon

Honorable
May 13, 2014
18
0
10,510
This computer has had a small history of malfunctioning, and I can't figure out what's wrong.

CPU: AMD FX-8350 Black Edition
GPU: EVGA GeForce GTX 660 3GB
PSU: Corsair Builder 600W 80+ Bronze Certified ATX Power Supply
Mobo: ASUS M5A 78L-M/USB3
Mobo2: ASRock 990FX Extreme3


It all started some time in spring, I can't remember exactly when. My decently powerful computer was for some reason being stressed by Borderands 2. The monitor kept going black and then "nvidia display driver stopped responding and has recovered."

I thought this was curious, but thought nothing of it. Swapped to lower settings and decided I would figure out another time.

A few weeks later the error started happening at random times when just browsing the internet. Then, suddenly the computer would randomly start restarting or bluescreening.

2 days later, my motherboard died. Or rather, it became extremely crippled. It wouldn't post half the time, and when it did it would only work if I turned off all cores but Core 1 on the CPU. For all intents and purposes it was worthless to me. I didn't buy an 8 core processor to use 1 core.

I sent the CPU in to AMD, they said it was perfectly fine, but sent me a new one anyway. Same problem.

I sent in the Mobo several times to ASUS but they never fixed the problem. Eventually, I bought a new Mobo (ASRock), tired of ASUS's poor customer service. And all was well and dandy for a few months.

In fact, I was able to keep multiple games running at the same time, my favorite combo being League of Legends and Civ 5, since Civ 5 is turn-based. Neither are very taxing on my GPU and barely use it.

What happens next is the most curious of all, as it's the one thing that makes me wonder "is it my GPU or PSU"

My 2nd monitor blew in the middle of playing League of Legends (Civ 5 may or may not have been running, but I don't think so), far from a game that taxes the GPU. No warning. No oddities before that. Just a puff of smoke from the top of the monitor and it was dead.

Monitors aren't connected to the PSU. Only the GPU and the wall. And the monitor in question was not old. Only 4 years old, if even that much.

But the computer itself still ran fine for a few months. And I could still play League and Civ at the same time...That is until recently. The "nvidia display driver stopped responding and has recovered." has come back any time I try to run both at the same time. Thankfully it only happens so far in that exact instance, but still, if I don't fix this problem fast, I may get more dead hardware.



Sorry if this story was a little long with unnecessary detail, but I'm just going inane over this stupid computer.

TL;DR:
"nvidia display driver stopped responding and has recovered." when trying to run games the computer is perfectly capable of running.
sudden crashes and bluescreens
mobo dies
new mobo, things work great again
monitor dies
"nvidia display driver stopped responding and has recovered." when trying to run games the computer is perfectly capable of running.


Additional info:
Estimated Wattage: 363, so I know my PSU isn't just being overtaxed.
I clean my computer regularly. Neither the GPU or CPU is overheating.
Just ran Furmarks and it's NOT causing the error, even though that's designed to keep your GPU running at 100% the whole time. I don't even =_=;
 
Solution
The 2 most common causes of your, Nvidia Display error is a overclock that is not stable.(Too much overclock) and not enough voltage.
Or low voltage drops on the 12 volt rail.
2 things you did not list.
Which model of Corsair power supply. They have very good models and so-so models. The CX series has 40c caps and if it overheats voltages get out of spec.
Case model. Older cases with top mounted power supplies used the power supply as the main exhaust. Which was fine because older CPUs and GPUs did not create as much heat.
Try running prime 95 and MSI Kombuster at the same time to see if you can replicate the problem. You can minimize both and they will still run.
While running check voltages and temperatures with Hardware monitor.
The 2 most common causes of your, Nvidia Display error is a overclock that is not stable.(Too much overclock) and not enough voltage.
Or low voltage drops on the 12 volt rail.
2 things you did not list.
Which model of Corsair power supply. They have very good models and so-so models. The CX series has 40c caps and if it overheats voltages get out of spec.
Case model. Older cases with top mounted power supplies used the power supply as the main exhaust. Which was fine because older CPUs and GPUs did not create as much heat.
Try running prime 95 and MSI Kombuster at the same time to see if you can replicate the problem. You can minimize both and they will still run.
While running check voltages and temperatures with Hardware monitor.
 
Solution