PC shuts down while gaming

[2Fast2Quik]

3 minutes ago, Hunched said:

HDD passed tests.

I'm in safe mode, a bit of the unfortunate thing is I can't even control my fans or monitor temps...

Soooo hopefully no new issues happen like overheating. 

I really wish I could monitor temps, voltages etc in safe mode, this is very limiting.

 

Keep me updated! =)

[Hunched]

27 minutes ago, 2Fast2Quik said:

Keep me updated! =)

Well Witcher 3 runs at 0fps and takes 20x longer to load.

Basically 0 stress being placed on the GPU, CPU, PSU, RAM.... sooo this is pretty fucking useless, I can't stress anything in safe mode.

I don't know how in the fuck I'm supposed to easily discover whether or not this is a hardware or driver issue when I can't test them separately.

Driver Verifier is supposed to be the best way to find bad drivers, but I think it's broken for me, it accuses everything, and when I fix everything it accuses, it still refuses to boot into Windows when enabled.

 

It's impossible for me to get into Witcher 3 with Driver Verifier enabled, which is kinda what I need to do to see if it's a driver causing the crash in Witcher 3.

But I can't even get into Windows with it enabled for some reason. I see lots of people who just get stuck in endless BSOD loops with Driver Verifier on Windows 10.

 

At least it's not the HDD.... Yay... 

If I'm stressing RAM I'll need to do it overnight with MemTest86. 

I'm going to DDU my Nvidia drivers and clean reinstall. It's rare that they would cause a sudden reboot but nothing's impossible. 

 

I wish Driver Verifier worked, maybe I'll give it another go, but it seems retarded. "This is the problem! No this is the problem! Now this is the problem! Oh shit you uninstalled all the 3rd party drivers I said were problems, well fuck you, here's a black screen you're still not getting into Windows" - Driver Verifier 2016

I really feel I need to get it working properly to solve my issue because this shouldn't be happening, I don't know why it's so broken.

I can't find anything saying how to fix it so you can get into Windows with it enabled. 

[2Fast2Quik]

2 minutes ago, Hunched said:

Well Witcher 3 runs at 0fps and takes 20x longer to load.

Basically 0 stress being placed on the GPU, CPU, PSU, RAM.... sooo this is pretty fucking useless, I can't stress anything in safe mode.

I don't know how in the fuck I'm supposed to easily discover whether or not this is a hardware or driver issue when I can't test them separately.

Driver Verifier is supposed to be the best way to find bad drivers, but I think it's broken for me, it accuses everything, and when I fix everything it accuses, it still refuses to boot into Windows when enabled.

 

It's impossible for me to get into Witcher 3 with Driver Verifier enabled, which is kinda what I need to do to see if it's a driver causing the crash in Witcher 3.

But I can't even get into Windows with it enabled for some reason. I see lots of people who just get stuck in endless BSOD loops with Driver Verifier on Windows 10.

 

At least it's not the HDD.... Yay... 

If I'm stressing RAM I'll need to do it overnight with MemTest86. 

I'm going to DDU my Nvidia drivers and clean reinstall. It's rare that they would cause a sudden reboot but nothing's impossible. 

 

I wish Driver Verifier worked, maybe I'll give it another go, but it seems retarded. "This is the problem! No this is the problem! Now this is the problem! Oh shit you uninstalled all the 3rd party drivers I said were problems, well fuck you, here's a black screen you're still not getting into Windows" - Driver Verifier 2016

I really feel I need to get it working properly to solve my issue because this shouldn't be happening, I don't know why it's so broken.

I can't find anything saying how to fix it so you can get into Windows with it enabled. 

 

Reinstall gpu as well ok. 

[Matt420740]

A. What are your temps when running Witcher prior to shutdown?

 

B. Do you have any other hardware intensive games to run to see if the problem is isolated to Witcher 3?  Figuring that out will help narrow things down quite a bit.  

 

You should really attempt to locate the issue first before you continue shooting in the blind messing with around with drivers.

[Hunched]

1 hour ago, Matt420740 said:

A. What are your temps when running Witcher prior to shutdown?

 

B. Do you have any other hardware intensive games to run to see if the problem is isolated to Witcher 3?  Figuring that out will help narrow things down quite a bit.  

 

You should really attempt to locate the issue first before you continue shooting in the blind messing with around with drivers.

Temps are like mid 60's

I just ran Unigine Valley and my PC is rebooting in it too after a short while. So it seems like my GPU or PSU is fucked, or the cable between them?

My PSU is only a month old and has been working fine, no issues ever at idle, didn't have issues under load until this. 

It's a Platinum rated EVGA 650 P2.

 

Temps and voltages are fine. I'm so fucking frustrated. I've spent the past 2 days trying to figure this out and haven't gotten much of anywhere.

It could still be a software or driver issue of some sort, I've seen so many topics where someone buys a new PSU and the same fucking issue still happens to them.

 

Fuck everything right now. I just want this to be over with already, I have enough shit to deal with as is, this is making me feel very defeated.

 

Having my PC constantly fucking shut down isn't good for anything, just makes everything worse each time.

I don't get a BSOD, I have no error logs or anything to figure out shit. 

 

There doesn't seem to be a way to even rule out drivers, safe mode is useless for testing.

I don't have spare components laying around to swap in to test.

I'm at a fucking loss.

 

I don't even know what the fuck to try anymore. Do I just start slowly rebuying my whole PC piece by piece with what I think is broken first until it stops?

That's a stupid choice but I'm not getting anywhere.

[Hunched]

GPU issues aren't even supposed to be able to reboot your PC, I get no artifacting, the GPU intensive apps don't crash, drivers don't crash.

PSU could be faulty, power cable could be faulty, GPU could be faulty, some driver could somehow be seriously fucking with things.

 

[Matt420740]

9 minutes ago, Hunched said:

I don't even know what the fuck to try anymore. Do I just start slowly rebuying my whole PC piece by piece with what I think is broken first until it stops?

That's a stupid choice but I'm not getting anywhere.

 

Calm down man.  We'll narrow it down and get it figured out.  You overclocking anything?  What are you monitoring temps with?  In your first post you mention you did a system restore prior to the issue.  Did the problem start immediately after the system restore?

 

 

[Hunched]

4 minutes ago, Matt420740 said:

 

Calm down man.  We'll narrow it down and get it figured out.  You overclocking anything?  What are you monitoring temps with?  In your first post you mention you did a system restore prior to the issue.  Did the problem start immediately after the system restore?

 

 

When I say I did a system restore prior to it happening, I mean I restored to a point in time before the issues existed, and they still happen.

The issues are happening with the overclock removed. I'm monitoring with GPU-Z and logging to file.

 

The issue happened overnight, when I last touched it, it was fine, come back to how it was when I left it, suddenly new problems exist.

I have no explanation as to why this might have started happening, and my PC isn't giving me dump files or anything I could possibly use to diagnose the issue.

I beat the entire Witcher 3 Blood & Wine expansion without much issue. It crashed to desktop once, maybe twice. 

That's about all I have to go on, it's all that happened which wasn't good, but I chalked it up to the game.

If that issue is the issue I'm having now, it went from rare crashes to desktop to PC reboots without warning every 30 minutes under load real fast.

 

[Hunched]

It just crashed while fucking underclocked. It's like 300mhz less on core and mem than I was running previously.

I'm leaning towards PSU at this point, but how in the fuck does one test that without specialized equipment?

It's one of the best PSU's money can buy, it's what I'm going to buy if I buy another PSU even. It's super quiet, good price, high reputation. Maybe I just got real unlucky.

 

If it's the PSU, it conveniently took a while to start acting up. It might be too late to RMA to the retailer at this rate.

It feels pretty toasty on the exhaust but I'm not sure how it felt when things are normal, I didn't inspect how it was when things were fine obviously.

I've also had 10 unexpected restarts in the past 24 hours, this isn't good for any of the components... I hope my PSU isn't damaging everything else if this is the case.

 

I can't see the PSU fan but I can feel the breeze underneath with my hand.

For what it's worth, the majority of the cases where shit like this happens seemed to be resolved by PSU replacement. 

 

It seems like the PSU might be overheating and shutting off it anything.

[Matt420740]

3 minutes ago, Hunched said:

When I say I did a system restore prior to it happening, I mean I restored to a point in time before the issues existed, and they still happen.

The issues are happening with the overclock removed. I'm monitoring with GPU-Z and logging to file.

System restore rarely fixes any real issues, as its pretty limited in what it can do.   It won't even get rid of a virus / malware.

 

Use DDU to uninstall the GPU drivers -  http://www.wagnardmobile.com/

 

Then do a clean install of the latest drivers from Nvidia - http://www.geforce.com/drivers

 

If the issue still persists after that try HWMonitor to watch the voltages and temps in real time while gaming in windowed mode to see if you see any issues.

 

I don't see anything pointing to your PSU being at fault here yet.  Especially if the voltages appear fine in HWMonitor.

 

 

 

[Hunched]

11 minutes ago, Matt420740 said:

System restore rarely fixes any real issues, as its pretty limited in what it can do.   It won't even get rid of a virus / malware.

 

Use DDU to uninstall the GPU drivers -  http://www.wagnardmobile.com/

 

Then do a clean install of the latest drivers from Nvidia - http://www.geforce.com/drivers

 

If the issue still persists after that try HWMonitor to watch the voltages and temps in real time while gaming in windowed mode to see if you see any issues.

 

I don't see anything pointing to your PSU being at fault here yet.  Especially if the voltages appear fine in HWMonitor.

 

 

 

I already did DDU and clean install, just a few hours ago.

I've been running the benchmarks windowed with GPU-Z over top and everything is solid with the GPU, at least it appears to be.

 

I don't think the PSU is failing to provide stable voltage, at this point if it's a PSU issues it looks like it might be overheating somehow and quickly shutting off.

I feel like the back of it felt warmer than it should have but I really don't know.

 

The PSU is about 25 days old, but it's the only component I have added/replaced in my PC in the past year.

It's supposed to produce very little heat since it's platinum rated and all. 

It's just barely under warranty (30 days) from NCIX so I may just do that and hopefully get a free replacement that solves everything.

At least if it doesn't solve everything, it rules it out just at the cost of shipping I guess. Better than having to buy a whole new GPU.

 

I'm going to inspect it a bit more, make sure the fan is running when it's under load, that nothing is somehow blocking it.

I have dust filters and everything so it should be fine... but everything else should be fine too.

There isn't much of a good reason for me to think anything would be broken.

[Hunched]

Well FUCK apparently NCIX only covers it for 15 days.

So I would have to go through EVGA and deal with costly and timely shipping to the US.

That idea just became a whole lot shittier, I was certain they did 30 days on everything.

[Matt420740]

9 minutes ago, Hunched said:

Well FUCK apparently NCIX only covers it for 15 days.

So I would have to go through EVGA and deal with costly and timely shipping to the US.

That idea just became a whole lot shittier, I was certain they did 30 days on everything.

I don't even know why you want to RMA the PSU anyways.  Nothing you've said here in this thread points to a failing PSU. 

 

What you need to do is actually find the issue before you go wasting time RMA'ing anything.   

 

Run a hard stress test on the CPU.  Does it shut down? 

 

If no do the same to the GPU.  Does it shut down? 

 

If so, you've found your problem, if not, we can keep going until we do find the problem.  Or you could just RMA shit at random until you figure it out........

[Hunched]

15 minutes ago, Matt420740 said:

I don't even know why you want to RMA the PSU anyways.  Nothing you've said here in this thread points to a failing PSU. 

 

What you need to do is actually find the issue before you go wasting time RMA'ing anything.   

 

Run a hard stress test on the CPU.  Does it shut down? 

 

If no do the same to the GPU.  Does it shut down? 

 

If so, you've found your problem, if not, we can keep going until we do find the problem.  Or you could just RMA shit at random until you figure it out........

The GPU benchmarks make my PC shut down.

I notice the PSU is quite toasty, and looking at the dust filter for the PSU fan intake, it's 100% clean. I'm not sure the fan has ever even turned on.

It's fine if the fan never turns on, provided it's cool enough. I'm not sure though.

 

The GPU isn't overheating, the PSU could be overheating though?

Maybe my PSU fan has been broken since day 1 and it just hasn't been hot enough in my room for it to overheat until now...

 

Keep in mind when the GPU/whole system is under load, that puts load on the PSU. Witcher 3 would be a very demanding game on all components, requiring a lot of power.

I don't doubt the build quality of the PSU overall, but if it's simply getting too hot... maybe that's a possibility?

 

We can't really know anything for complete certain unless I had a spare GPU or PSU to swap in and see if the issue persists as far as I know

[2Fast2Quik]

26 minutes ago, Hunched said:

Well FUCK apparently NCIX only covers it for 15 days.

So I would have to go through EVGA and deal with costly and timely shipping to the US.

That idea just became a whole lot shittier, I was certain they did 30 days on everything.

 

Hunched have you reinstalled the gpu yet, I've had something very similar happen to me and that was the solution, please tell me you have.

[Hunched]

For anyone who may stumble upon this in the future, here is the continuation future being!

https://linustechtips.com/main/topic/607007-how-do-i-test-my-psu-for-failure/#comment-7860141

 

[2Fast2Quik]

11 hours ago, Hunched said:

For anyone who may stumble upon this in the future, here is the continuation future being!

https://linustechtips.com/main/topic/607007-how-do-i-test-my-psu-for-failure/#comment-7860141

 

I salute you for adding that to the thread, help out a poor chap in the future, good for you man, and congrats on your functioning pc now.