GPU Crashing Frequently In Gaming [Faulty GPU?]

BlackBird337

Honorable
Sep 12, 2013
13
0
10,510
Hello, let me start with my system specs.

Intel Core i7-4790K CPU @ 4.00GHz
Threading 1 CPU - 4 Cores - 8 Threads
Frequency 4300.54 MHz (43 * 99.98 MHz)
Architecture Haswell (22 nm)
CPUID / Stepping 6.C.3 / 6.3C (C0)
CPU Extensions MMX, SSE, EM64T, VT-x, AES, AVX, AVX2, FMA3
Caches L1 : 128 KB / L2 : 1024 KB / L3 : 8192 KB
Microcode 0x000001C
Max. TDP 88 Watts
Temperature 47°C
Voltage 0.128 Volts
Type Stock
Motherboard
ModelMSI Z97 GAMING 7 (MS-7916)
BIOS American Megatrends Inc. - V1.8 - 12/24/2014
CPU Socket Socket 1150 LGA
North Bridge Intel Haswell rev 06
South Bridge Intel Z97 rev 00
Memory (RAM)
Size8192 MB
Type DDR3-1333 - Dual Channel
Frequency 666.8 MHz (1:5)
Timings 9-9-9-24
Slot #1 ModuleA-Data Technology - 4096 MB (10700)
Slot #2 ModuleKingston - 4096 MB (10700)
Graphic Card (GPU)
GPU TypeNVIDIA GeForce GTX 660 (549 MHz) [[[USING LATEST DRIVERS AS OF 7/6/2016]]]
VRAM Size 2048 MB (3004 MHz)

PSU: Modular Corsair 600w..don't remember the rest of the details on it


Now to the problem..

I've been having this problem for a very very long time so far, and it is overall a hit and miss scenario. Sometimes my GPU will just stop and sometimes it wont when gaming. Before my GPU, during gaming takes a huge dump a few things occur first: Sounds in my game, or anywhere you will hear clicking in the sound (clicking being the sound 'flickers' or something), next video begins to slowly lag upon these flickers. When the GPU is ready and call it a day the geometry/world geometry in games would begin to flicker and change about, sometimes the world flickers entirely. I had one instance where I was playing TF2 and when looking around my gpu freezes, I was able to still look around for a split second but noticed a window didn't move with everything else. Recently I've played Dead by daylight and then when my gpu broke I saw the characters really deform and hilarious looking, almost something out of Garry's Mod.

After the crash occurs I get greeted by the kernel failing on my gpu and having to restart. If I keep gaming the problem persists and it will keep crashing until the gpu begins to slow down my computer entirely, so I just restart my computer and 6/8 times I run everything happy gilmore.

My temps don't seem absurd when these crashes occur, they are about at 70-90c when under load.

I bought a PSU tester and everything was green and lit up and all the voltages and rails were right on par with its specifications.

Today [7/6/16] when playing overwatch my game was just lagging so horribly with geometry flickering until my gpu gave up and the image froze, but the sound didn't really freeze..from there my computer just simply shut down and restarted...first time it's done that ever. I restarted my computer and everything seems to run okay for the time being.

When playing low gpu intensive it wouldn't ever crash unless its TF2. Gmod runs well without any issues for long periods of time.

I recently bought a GTX 960 and it will arrive on 7/12/2016 as I'm thinking my GPU is faulty...but is it really? Does my GTX 660 met its match and should be put down? Could I RMA it for a new one then resell it?

I could use some help on this. Thanks.


EDIT: Forgot one more thing. I checked my Memory.dmp for when my gpu shut off my computer and it showed this:

*******************************************************************************
* *
* Bugcheck Analysis *
* *
*******************************************************************************

Use !analyze -v to get detailed debugging information.

BugCheck 116, {fffffa800bb2a010, fffff880087caa08, ffffffffc000009a, 4}

*** ERROR: Module load completed but symbols could not be loaded for nvlddmkm.sys
Probably caused by : nvlddmkm.sys ( nvlddmkm+7aca08 )

Followup: MachineOwner


EDIT2: Also I've reinstalled drivers more than a million times, even during safe mode. I've used a program to remove drivers completely and clean it out of my system but no dice.
 
Sorry, but I do have to sleep tonight. I cannot take the time to read all of that..

Download DDU (Display Driver Uninstaller) by clicking on the icon below, and follow the prompts to get into safe mode, and then choose the top option there. That will clean out all parts and pieces of any previous video drivers that have ever been on your system.

Once that is done, bppt back to the Desktop, Download the latest drivers, and install them.




 

BlackBird337

Honorable
Sep 12, 2013
13
0
10,510


This has been done twice so far.
 
You have non-matching memory sticks. Try removing one of them, and see if the system stabilizes. I know running with 4GB is going to make things slow, but if the system stabilizes, then we probably have identified the problem. If you get lots of errors with 1 stick, swap memory sticks and try everything with the other stick.

Lets see if that changes anything at all. Please let me know what happens.
 

BlackBird337

Honorable
Sep 12, 2013
13
0
10,510
Sorry for the delay..soo..Alright.. despite having non-matching sticks they run at the same frequency and don't cause issues.

However this is aside from my point. I recently bought a new gpu an MSI GTX 960 4GB since I was planning on anyways. Everything ran very well. However the issue occurs but doesn't crash my driver. So I did some digging and found the solution to the problem.

The Killer Ethernet e2200 driver from MSI was causing a memory leak, causing my non-paged memory pool constantly going up and reaching to 4+gb of memory. Due to this, my computer has been constantly using 6.0-7.0GB of RAM memory when the computer was idle and when ingame and I'm reaching that ram limit, my computer would begin to fumble and act up badly as explained earlier. Upon changing the registry code for the driver to 4 which disables it, my issue has been resolved and it doesn't seem like my GPU was faulty after all meaning I can sell it for some extra cash!