Series of odd problems, unsure of culprit.

[ShadowScotsman]

Hi there ive been strugling with a series of problems over the last month and im nearly out of ideas as to what is causing them. Im hopeing some of you guys might be able to shed some light on the matter as the only options im thinking of next are considering parts faulty(which doesnt make a whole lot of sense to me for reasons ill go into below). This all started when i bought a new graphics card and ram 1 month ago, a gtx 1080 and 16gb of ram (4 x 4gb) i later realised my cpu at the time wasnt quite up to running vr games and so i then ordered a new cpu and motherboard, the cpu and motherboard arived last week, since installing everything ive been lucky if i can watch youtube consistantly, playing games is prety much impossible (crashing or hanging after anything from 5 to 30 mins of gameplay)

Ok so there have been a few different problems, some that i think ive solved and others that are still persisting, ill list them in the order the started happening. Along with my PC Specs.

PSU: Cooler Master G650M
GPU: Nvidia GTX 1080
RAM: 4x 4GB Team Group Elite Black DDR3 1600MHz Dual Channel
HDD: Western Digital WD10EZEX 1TB
MB: ga-970a-ds3p fx (rev2.1)
CPU: AMD fx 8370 (witch bundled wraith cooler)

Not sure if its important or not but the things plugged into the pc are, a keyboard and a mouse both using usb ports on the motherboard, my ethernet cable again pluged into the motherboard and two monitors, one pluged into the hdmi port on the 1080 and one on the dvi-d port using a dvi-d to hdmi cable.

After getting the new gtx 1080 and ram and installing it (1 month ago):
Was unable to play games, game would hang/freeze and event viewer show that nvlddmkm had stopped working and succesfully recovered.
Sometimes while turning the pc on in the morning or after a restart it would hang on the bios splash screen, or hang on a black screen while giving a beep code of infinite short beeps (a code that doesnt exist according to my old motherboard information)
Switching the keyboard and mouse to different usb slots on the motherboard solved the bios hang and beep code problem although i do not know why.

Once the new motherboard and cpu arived i installed the new hardware only to find the nvlddmkm stopped working error was still around and seemingly got worse.
Now it would also give that error while watching youtube sometimes (it was begining to seem like anytime the pc was doing something graphicly intensive it would give that error)
i tried uninstalling and reinstalling the nvidia drivers multiple times but to no avail, enfact at one point it caused windows to hang on a black screen for multiple minutes during startup although i later discovered this was due to a driver conflict(windows installed the standard vga graphics drivers before i could install nvidia driver meaning i had to force windows to not install drivers)

During my multiple attempts at fixes one very unusual problem occured, i had the pc running (able to watch youtube with ocasional display driver resets) i headed off for bed that night shutting down the pc and the next morning it wouldnt load to windows, just a black screen after the bios splash and then the computer would auto restart and just keep looping with that happening, eventually while trying switching hard drives, graphics cards ect i checked the bios and changed one setting on a whim, i changed how it was treating the sata connection, i do not remember what its defualt was but it was at that point on the setting AHCI, the other options were RAID and NATIVE IDE, i set it to NATIVE IDE, and now the hard drive would load although windows was corrupt in some way, it did the "Starting windows" with black screen and the logo would get about half way through its little animation before freezeing and BSOD, the BSOD moved to fast for me to read but ultimately one more reformat and reinstallation solved that issue i think. im not sure if that setting was always on AHCI or if it changed itself somehow so im at a loss as to what caused that one.

And now im back to square one with no real sollution in sight, the nvlddmkm stopped working error keeps happening although i know this is actually the windows tdr, but as to what is causing windows tdr to reset the driver i dont know.

Here are the tests/solutions i have tried so far.

Hard Drive:
Have reformatted multiple times with fresh installs of windows.
Checked the drive for errors but it came back clean on a short and long generic test.
Have used a different old hard drive, again reformatting and installing windows fresh.

Ram:
Ran a pass of memtest from a usb, completed 1 cycle with no errors.
Swapped old the new Ram for the old Ram i was using before the new gtx 18080.
Double checked that the timing in bios are set correctly for my ram.

PSU:
Checked using multiple online calculators that my psu is up to the task of running my parts.
Checked in bios that each rail was reading the correct voltage.

Graphics Card:
Have checked multiple times that the card is correctly seated in the pci-e x16 slot.
Checked it is in the primary pci-e slot according to the motherboard manual.
Set power management mode to prefer maximum performance in nvidia control panel.
Swapped out the new card for my old card (a radeon hd 7790) had the same problems but now instead of nvlddmkm it was the amd variant or it, it was something like amdddmkm i cant remember exactly.

Motherboard and CPU:
I am unsure of why they would be the culprit as the problem arose well before those new parts were installed.
And unfortunately i belive the old motherboard and cpu is dead. i tried to detach the fan/heatsink from the old cpu but the thermal paste had glued the parts together resulting in me pulling the heatsink with cpu attached right out of the socket, so yea fairly sure im not able to swap those parts back to test.


So im now at a point where i feel like ive tested everything i possibly can, and im not entirely sure what the best course of action from here is. Id grately appreciate any advice anyone could give in this situation as this is my first time building/upgrading the pc myself, origonally the pc was bought as a completed unit and i simply had to buy and install windows myself.

 

[jsmoke]

I feel like Hercule Poirot. It was Col. Mustard with the Graphics card in the study...

Is the Ram not the only common factor, I.e you've tried different GPUs so could sort of rule out the GPU and it happened before new motherboard. Try with just one stick of ram in various slots and then another stick etc.

 

[ShadowScotsman]

The only common factors between before and after installing the new motherboard are the graphics card, ram, hard drive and psu, although the problem wasnt there before i added the graphics card and ram so i dont see it being the psu or hard drive unless they have coincidentally failed at just the right time.

i currently have 6 sticks of ram in total, 2x 4GB kingston and 4x 4GB Team group the kingston was the origonal ram i had before getting the graphics card and the new team group ram, i have tried taking the new team group ram out entirely and reinstalling the 2 old kingston sticks and it had no effect at all. even made sure i was putting the kingston sticks in the right slots for dual channel according to the motherboard manual i.e
Emtpy slot
4GB Kingston
Empty slot
4GB Kingston

thats what makes no sense to me about this problem, it seems like ive swapped out all parts that could be the culprit yet the problem persists, originally it had me thinking it must be a software issue but as ive had problems that have happened during boot and bios that and ive done fresh installs and reformats that to me says it cant be software related.

hmm thinking about the psu might be a common factor, my old graphics card had the fan on it die, since i was getting a new card i simply avoided any gaming on it since i knew it would overheat so in theory its possible the issue existed from the psu (about 3-4 months ago) and ive simply not been doing anything graphicly intensive enough to trigger the problem until now. i do still have my old psu its only a 500w and lacks the pci-e cables to run the 1080.

thinking about it i should be able to test if its the psu by confirming my old graphics card causes the same issues and problems, then switching the psu to the old 500w one and checking for the problem again.

ill try that test to see if the psu is at fault and if its not the psu then ill try checking the ram one stick at a time like you mentioned. im just hopeing the pc survives until tomorow when i can do those tests.

 

[jsmoke]

It is a bit weird, a bad stick of ram would probably be consistent with the symptoms though.

 

[Blackjack Davy]

Which display driver version are you using? I've had endless problems with black screens, BSOD's, "Driver has recovered from an unexpected error" messages on my 1080ti. Absolute hell. Just like you swapped out just about every piece of hardware.

Problem was worst with Gsync and windowed mode that was a guaranteed crash, fullscreen was ok. Someone recommended driver version 397.64 WHQL and its been rock solid since. I know theres a newer one available but I'm reluctant to risk it.

 

[EsaT]

If previous graphics card was AMD then changing to Nvidia without full removing of drivers could have caused some conflicts, but OS reinstall not doing anything scraps that.
AHCI/IDE setting would only cause problems in booting into Windows.

Also problems in BIOs phase of boot hints toward hardware problem.

That PSU certainly isn't strong candidate for longevity.
Finest Chinese crap from Elite as primary caps.
And then some of secondary/smaller capacitors from Jun Fu, because at best third tier C(r)apXons weren't cheap enough to put them everywhere in secondary.

 

[Journey]

Is your new motherboards BIOS up to date, and was your old one for that matter?

 

[ShadowScotsman]

i tried alternating the ram like you suggested and it didnt have any real effect, other than it taking a little longer to boot up, but the driver crash during games still happened in all configurations of ram i tried.

i tried switching the psu but as mentioned i can only check that with the old graphics card and with the dead fan on the old card i cant run games at all really without it overheating and forcing a shutdown, i didnt have it crash while browsing youtube but its entirely possible i just got lucky and the crash didnt happen for the few hours of testing i did with it.

ive tried using the latest 4 driver versions with no luck, some slight changes in the speed the error happens but overall the problem persisted.

i doubt its a display driver overlap/conflict, i had that happen between the windows standard vga driver and the nvidia driver at one point and since ive been using DDU to remove the old drivers and using gpedit to ensure windows doesnt install its ow drivers.

and i checked the bios of my new motherboard, unless im missing something it never got a new version released, checked the manufaturers site and the only bios version the have listed is F1, which is how the board was shipped. im guessing it being the fx version of that board and thus being revision 2.1 means it comes loaded with the last bios version released. and yea i had updated my old board bios about a year ago as part of troubleshooting a problem(turned out to be a dead cmos battery)

looking back through all my notes and whats been spoken about here im left thinking it has to either be the graphics card or the psu, and my only test for the psu means switching the graphics card aswell(since the old psu has no pci-e cables)

and worse yet is i was reading up about rma details and the warranties and it seems i would need the origonal packaging that the parts arived it, which i dont.

so unfortunately unless someone out there has had these exact same problems and knows the special fix i think im up the creek without a paddle.

Thank you for all your help though, i appreciate you all taking the time to try and help me.

 

[Finners]

Sorry if I missed it but what was the GPU before your new 1080?

Always a mine field buying second hand parts which I assume you have done with that CPU. Was the 1080 new or second hand?

 

[ShadowScotsman]

all of the parts have been new, wanted to avoid the possibility of getting used parts that might be bad.

the old graphics card is a radeon hd 7790, the fan on it is dead though, i think its just gathered dust over time and got gummed up.