Hi!

So recently I have started having issues with games on my rig. Basically, whenever I play something a bit more intensive (eg. Overwatch on high settings, AC: Odyssey, Darksiders Genesis (?) ) my computer will either crash the game to a black screen with audio or no audio at all or restart completely.
I have kept an eye on temps, my aircooled 2700x (No OC) sits at most at 65C under very heavy load.
My GPU can get pretty hot (75-80 C) but as far as I've researched it's not too bad for an aircooled 2080ti.

Things I have tried so far:

Reinstalling Windows (this copy is about 10 days old)

Reinstalling GPU drivers with DDU in safe mode

Running games off a SSD instead of a HDD (someone suggested it might be a loading time issue).

Adding extra fans to my system (currently 7 fans configured as 5x intake and 2 as exhaust).

A plethora of other fixes and tips suggested of reddit, LTT, TH etc including fiddling with power settings and plans, disconnecting my secondary monitor and so on.

Now I've had an EVGA 750w G PSU fail on me last year and I eventually bought a Corsair HX850i in the month I was waiting on the RMA for the PSU.

The PSU is literally the only component I haven't swapped so everything seems to point at that, I find it hard to believe that I am THAT unlucky to have TWO PSUs from "reputable" manufacturers fail on within the space of one year but I guess stranger things have happened.

So what do you guys think? Any suggestions?

Specs:

Mobo: Asus Crosshair VII HERO (WIFI)

Ryzen 2700x (no oc) on Noctua NH D15

TridentZ 16GB RBG RAM, 3466Mhz DOCP

EVGA 2080ti XC (no oc)

OS on Samsung Evo 970 M.2 500GB SSD; Games running off a Samsung Evo 860 1TB; 2 TB WD Black as extra storage.

PSU: Corsair HX850i

3x Noctua 120mm iPPC 2000 fans, 3x Corsair ML120, 1x Noctua Chromax 140mm

Case: Fractal Design R6

You could try to repaste the unit with higher grade thermal compound. Does not void the warranty, but in my experience, does help to a certain degree. If you can, post your system photos as sometimes poor airflow can cause a rise in temps.

Additional info:

Things I have tried (some again) in the meantime:
A plethora of benchmarking/stress test tools:

Furmark, Heaven, Valley, Prime95, Memtest86, 3DMark Fire Strike and Time Spy, Windows Memory Diagnostic Tool.

Under extreme synthetic load (tried the 4K bechmarks and stress tests) the highest I've seen is 67C on the CPU and 84C on the GPU.
I ran Time Spy and Heaven for 3 hours each with no glitches, crashes or any issues.
I read somewhere that certain services like Discord or Logitech Hub might make the problem worse by crashing the Nvidia drivers...I will have to test for these as well.
I will try the new Nvidia drivers tonight after a DDU cleanup of the old ones.
Someone said I need a 1600W PSU? Like, I don't feel my system draws that much but idk, what do you guys think?

transdogmifier
What motherboard are you using? Did I miss it? Have you tried updating the
BIOS on it?
 
Believe it or not it can affect things...some bios don't have voltages set right or
something screwy is in them in other ways.
 
 

Oh, my bad I forgot to mention it. Asus Crosshair VII Hero (Wifi). I have updated to the latest BIOS (3004).

spit051261
Everything you have done and have seem fine .
Need to try GPU in another computer to see if you get the same problems.
Don't think your GPU paste is at fault .
Something more sinister.
One other thing , asus boards are not that good dealing with fast ram and a ram fault will run you around in circles.
Not saying it's ram BTW.
Does your board show Q Codes ?

Well, I do have another computer but it's a very old AMD Phenom X4 that I've converted in a family media server of sorts (I can still play 4K video and browse, maybe some old school gaming too) that's got a 970 in it. My PSU return is in it so 750w let's hope it's enough.
I have tried fiddling with RAM before and Memtest seemed fine. I will try to drop it to 3200Mhz to match Ryzen's IF and see if that improves stability.

RaevThaonai
Additional info:

Things I have tried (some again) in the meantime:
A plethora of benchmarking/stress test tools:

Furmark, Heaven, Valley, Prime95, Memtest86, 3DMark Fire Strike and Time Spy, Windows Memory Diagnostic Tool.

Under extreme synthetic load (tried the 4K bechmarks and stress tests) the highest I've seen is 67C on the CPU and 84C on the GPU.
I ran Time Spy and Heaven for 3 hours each with no glitches, crashes or any issues.
I read somewhere that certain services like Discord or Logitech Hub might make the problem worse by crashing the Nvidia drivers...I will have to test for these as well.
I will try the new Nvidia drivers tonight after a DDU cleanup of the old ones.
Someone said I need a 1600W PSU? Like, I don't feel my system draws that much but idk, what do you guys think?

My 2080 Ti XC Gaming did 56C stable in Unigine Heaven at 100% power, max fans in 20C air conditioned room.  If you're breaking 70C in Unigine Heaven at 100% power you have either a very hot room, not enough fan speed or BAD TIM.

kevinc313

RaevThaonai
Additional info:

Things I have tried (some again) in the meantime:
A plethora of benchmarking/stress test tools:

Furmark, Heaven, Valley, Prime95, Memtest86, 3DMark Fire Strike and Time Spy, Windows Memory Diagnostic Tool.

Under extreme synthetic load (tried the 4K bechmarks and stress tests) the highest I've seen is 67C on the CPU and 84C on the GPU.
I ran Time Spy and Heaven for 3 hours each with no glitches, crashes or any issues.
I read somewhere that certain services like Discord or Logitech Hub might make the problem worse by crashing the Nvidia drivers...I will have to test for these as well.
I will try the new Nvidia drivers tonight after a DDU cleanup of the old ones.
Someone said I need a 1600W PSU? Like, I don't feel my system draws that much but idk, what do you guys think?

My 2080 Ti XC Gaming did 56C stable in Unigine Heaven with at 100% power, max fans in 20C air conditioned room.  If you're breaking 70C in Unigine Heaven at 100% power you have either a very hot room, not enough fan speed or BAD TIM.

Hmm, my room sits around 21-22 C but tbh the Define R6 kinda sucks for thermals. On the other hand, the card has been sitting at those temps pretty much since I've had it so...idk
Please bear in mind that mine is an XC (a.k.a two fans) not an XC Gaming. Idk if that make a any difference.

If your ram is not set up right , your pc will just crash for what seems like no reason and you could be doing anything at the time.
Have the same with my 4400 and have to take it down to 4200 with XMP to stop crashes.
My bios also likes to change the voltage for no apparent reason which doesn't help.
Ram faults are not good .
Lucky my board shows Q Codes and they always end in 5 which is ram related .

Might be fine for a while then crash
Took it down to 4200 and now no crashes
Asus board max supported ram is 4400 apparently but the bios does strange things at that speed .

spit051261
Might be fine for a while then crash
Took it down to 4200 and now no crashes
Asus board max supported ram is 4400 apparently but the bios does strange things at that speed .

Thanks for the suggestion! So far, so good. 3200Mhz seems to work far better, even if the RAM is not on the QVL (I could have sworn I've checked this when getting it...maybe I just got bamboozled by the part numbers...)

Unfortunately the GPU still sits in the lower 80s...
I haven't had a crash since the RAM tweak but I think I might need to try and repaste the GPU.
I've got some Kryonaut, do you think that would be suitable? If not, which paste. I'm off this Friday and if I need to get it done, might as well order stuff off Amazon.
Also, can you reuse the thermal pads or do I have to buy a set of those as well?
Sorry, I've no idea how this works

spit051261
Re paste sounds good .
One of the main men will be along shortly to answer that one .

Thanks for all the patience, man!
Sorry I'm asking all the questions but this is my first build so there's plenty of stuff that I need to learn.

RaevThaonai

spit051261
Re paste sounds good .
One of the main men will be along shortly to answer that one .

Thanks for all the patience, man!
Sorry I'm asking all the questions but this is my first build so there's plenty of stuff that I need to learn.

Everyone is still learning and the only stupid question is the one that hasn't been asked so ask a way.

I'm using hynix I know it doesn't play nice with my asus board at full speed.

RaevThaonai

kevinc313

RaevThaonai
Unfortunately the GPU still sits in the lower 80s...
I haven't had a crash since the RAM tweak but I think I might need to try and repaste the GPU.
I've got some Kryonaut, do you think that would be suitable? If not, which paste. I'm off this Friday and if I need to get it done, might as well order stuff off Amazon.
Also, can you reuse the thermal pads or do I have to buy a set of those as well?
Sorry, I've no idea how this works

Either paste is fine but if the MX-4 is more than a couple years old I'd say use the new stuff just to be safe.  Use the hybrid directions I posted above to remove the cooler, I'll repost them below.  Either spread the paste flat plus a rice size dot, or use an X pattern, or a good fat pea sized dot.
 

Checking the TIM isn't hard.  Just follow the directions for the hybrid kit about how to take apart the card.  You want zero bare spots, you're far better off having 100% paste coverage and a bit extra paste.  Spread the existing TIM flat with a straight plastic edge and add a rice sized dot, or re-apply with your pattern of choice at least pea size.  It's a bit annoying but really should be checked if things are kinda hot and otherwise setup well.

 
Hard to say on the ram, 3200CL16 on hynix CJR (likely your ram chips) is pretty typical, but keep an eye on it under load and watch out for random WHEA errors.  The F4-3200C16D-16GTZR kit does appear on the MB QVL, that's just a shade down in speed from your actual kit, so if you're running similar timings with 1.35V you should be good to go.
 

Well, I went ahead and did it...it was easier than I feared. Needless to say there were multiple spots uncovered by TIM and the one left was caked and dry. Some Kryonaut later, I reassembled everything with the only issue being the small black grill that goes over the fan connector (EVGA, guys, come on...) that didn't wanna go back into its place. Got it there eventually and put it all back together.
Now the trouble is, when I turned on the PC it made this squeaky noise. At first I thought it was because I moved one of the case fans but I stopped them all by hand and the squeaking was still there.
So I took out the GPU and booted again, the squeaking was gone.
Now I'm afraid to put it back in because if there's something wrong with it I don't wanna make it worse.
What should I do, guys?
I should have known this won't go down as swimmingly as I thought it would.
Also, I noticed a bit of what looks like glue on the heatpipes. I know I am being paranoid but...it's not fluid leaking from one of the heatpipes...is it?
I took some photos of before and after but for some reason I cannot upload them? It gets stuck at 65% when uploading

What kind of squeaking are we talking about here and when does it happen?  The two big noise makers on GPU's are fan rub (on cable or heatsink) and coil whine.