Windows 10 Randomly crashes and restarts while gaming

Toggle sidebar Toggle sidebar
Status
Not open for further replies.
C

ChrisWKuiper

Honorable
Apr 7, 2015
46
0
10,530
A week ago my desktop started randomly crashing to black screen while gaming, and then immediately restarting and acting like nothing happened, no error codes, no nothing. Event Viewer just shows the typical critical error which dosent help much. The power supply is in good health and has plenty of wattage, all of my other components are fine, motherboard, ram, gpu, cpu, nothing is overclocked. I checked temperatures and voltages of all components, nothing is out of the ordinary. I preformed stress tests on all components and a mem test on the ram and everything passed with flying colors, so I don't think its a hardware problem.

Fast forward to today...

I rebuilt the computer just yesterday using a new motherboard, cpu, and power supply. The only components carried over from the previous rig are two of the ram sticks and the hard drive with windows on it. The random crash and restarts are still happening. Now it can't possibly be a hardware problem right? The ram passed mem test....so it must just be windows 10 shitting itself? I have made sure all of my drivers and motherboard bios are up to date, so the only thing I can think of at this point is to reinstall windows. I would prefer not to do that though until I at least have some idea of why this is happening, otherwise it might happen again....

Any thoughts would be appreciated.
 
Sort by date Sort by votes
is windows giving any errors about the crash or is it just black screen - nothing - then bios post?

cause black screen reboots are typically power or temp related. ram crashes tend to be a bit more random (and usually include some sort of blue screen crash). borked versions of windows is possible... you could try a SFC to see if anything comes up troublesome.

 
Upvote 0 Downvote


No errors given. The computer will just shut off (looks like a power outage) and then restarts like nothing happened. It usually happens when moving between a graphics intensive thing and a non graphics intensive thing, like moving from league to chrome and them back to league. I found this thread of someone else having the exact same problem: http://www.tomshardware.com/forum/id-3583313/computer-randomly-restarts-playing-games-windows-fall-update.html?xtor=EREC-8889

However I have updated all of my divers so I dont think that will help me.
Temps are fine, and like I said it has done this on 2 sets of motherboards, power supply's, and cpu's .

What is SFC?
 
Upvote 0 Downvote


I did a completely fresh reinstall of windows. The only things I downloaded after the fresh install of windows was league to test if it was still happening. It's still happening. When I get home I am going to completely uninstall and reinstall the drivers and firmware for my 980ti.
 
Upvote 0 Downvote


might be a bad power/reset switch in the case.

still going to underline my first comment that black screen restarts are generally electrical or temp related. A bad power switch in the case would count as a power problem.
 
Upvote 0 Downvote


how could It be a bad power switch though? I have replaced the psu, cpu, motherboard, psu cables, tried with and without a batter backup. Is there anything else in the power chain I am missing?

I also just experienced a rapid succession crash, it crashed while I was in a game of league (computer ceased to function then restarted) and when I logged back into league it tries to put you back into the game like its supposed too, however the moment I loaded into the game, windows crashed again. Then it restarted and I logged back into league a second time, and this time when I loaded in everything was fine again. It is truly bizarre.
 
Upvote 0 Downvote
I am having the same problem since my last win 10 update on July 31st. I can play Frostpunk for about 3min before the system hard crashes to black and reboots. Sometimes it reboots twice (power LED coming on, then off, then on again). Temperatures vary at the time of the crash. Sometimes the system crashes at 60°C sometimes at 80°C for the CPU. Crashes do appear related to on-screen action, especially particle effects and in-game videos kill it.

The system is new (2 months old) and used to work without a hitch up to that point. I tried everything from underclocking to cooling the ambient air with dry ice, temperature does not appear to be the issue as there are no slow-downs or visible throttling prior to the crash. I updated all drivers - did nothing, then reverted the NVIDIA driver to the last known stable version - did nothing either.

In the maintenance monitor it is clearly visible that the problems started after a particular win 10 update, which also updated a ton of drivers:
Intel(R) Xeon(R) E3 -1200/1500 v5/v6th Gen Intel(R) Core(TM) PCIe Controller (x16) - 1901
Intel(R) 100 Series/C230 Series Chipset Family PCI Express Root Port #9 - A118
Intel(R) 100 Series/C230 Series Chipset Family PCI Express Root Port #9 - A112
Intel(R) Managment Engine Interface
Intel(R) HD Graphics 630
Intel(R) 100 Series/C230 Series Chipset Family PMC - A121
Intel(R) 100 Series/C230 Series Chipset Family LPC Controller (HM175) - A152
Intel(R) 100 Series/C230 Series Chipset Family PCI Express Root Port #5 - A114
Intel(R) 100 Series/C230 Series Chipset Family SMBus - A123
Intel(R) 100 Series/C230 Series Chipset Family Thermal subsystem - A131
Intel(R) 100 Series/C230 Series Chipset Family PCI Express Root Port #4 - A113
NVIDIA GeForce GTX 1060
...
and many more.

But I suspect the Intel drivers are the culprits. If I had to guess, I would say the power management got messed up seeing how the system hard crashes without stutter or slowdown (which I assume would be the case if temperature were the issue). I imagine somthing is going wrong when both the CPU and GPU draw power simultaneously.

This would explain why rebuilding the PC wouldn't work if Intel power management was the issue. Moving all components to a new rig with the faulty drivers in place would not change anything.

Any help is greatly appreciated. Unfortunately, due to all the tinkering and waiting it out I lost my system restore point from before the updates, so reverting is not an option.

My rig:
Acer V17 Nitro
i7 7700HQ
8GB RAM
GTX 1060
 
Upvote 0 Downvote


your computer case has a physical power switch in it; I was suggesting that switch might be failing. You can test this by unplugging the power switch from the motherboard and using a flathead screwdriver to bridge the pins to start the system, maybe get someone's help to hold it there, and do some tasks that usually crash your pc. if the pc doesn't reset and the problem goes away then it's the power switch in your computer case.

Other things to test (for high temps) would be to remove the side of the case, stick a fan in the opening on high (room fan) and see if the crashing continues, if it doesn't then you have a temp problem somewhere and we'll have to figure out where.



you have a laptop, separate series of issues (though temp would rocket to the top of my list for you)
 
Upvote 0 Downvote


I agree it's a different realm, but the problem seems familiar. I ruled out heat as the problem, it is definitely voltage. Simply launching an application that uses the GPU will cause a crash. FurMark crashes after about 2 minutes on lowest stress (GPU 65°, CPU 60°). Interestingly enough, if I pull the power cord and run on battery, the system will last about 10 min before crashing.

The win/driver update is most definitely at the heart of this, since everything worked well before that (GPU at 80° for 3h, while running Spotify, Chrome, etc in the background was no problem at all). My only regret is not having a restore point 🙁

 
Upvote 0 Downvote


After reading all of what you wrote, I think we are having the same issue. Do you know exactly what windows update it was that changed all of the drivers? I am going to find a list of the recent intel chipset and power management drivers and see if I can find an older version from before whatever that windows update was.
 
Upvote 0 Downvote
have you tried OCCT, the PSU test in particular? can you access crash dumps (WhoCrashed)? the windows update was rather minor and installed updates on a lot of windows add-ons (paint, calculator, etc.) the only gaming related update I can see is the xbox gaming overlay that got installed.
 
Upvote 0 Downvote

Please don't use furmark to test... anything. Furmark (once upon a time) was a popular stressing tool. However, it stressed nvidia gpus so badly it was destroying them causing fires (actual fires). As a result nvidia wrote into their drivers code to THROTTLE your nvidia gpu whenever someone runs furmark. Furmark no longer is a valid stability test for an nvidia gpu. If you're crashing when running furmark, then it's either due to the throttling in the driver, or just a driver issue on the whole.

AS for AMD gpus, I tell people never to run furmark on them as well. Why? because it set people's computers on fire (granted they were Nvidia gpus) due to insane stress. No one in their right mind should run such a program on their pc unless they like shortening the life of their parts.


As to your guys problems... lets eliminate nvidia graphic drivers from the situation.

reboot your pcs into safemode, run DDU (display driver uninstaller)
reboot your pc
install the latest nvidia driver, test, if the problem is still there, repeat the above steps, only try an older nvidia driver (maybe something from a month or two ago) and try again.
 
Upvote 0 Downvote


I ran WhoCrashed and it shows nothing because my computer is not writing crash dumps. Its not a blue screen, its a black screen and a restart, the computer just dies and then comes back to life so it dosent have time to do a crash dump.

I am pretty sure you are on to something though with some power management driver being the culprit. This issue ONLY happens right at the moment when the cpu and gpu are both asking for more power at the same time, like when just loading into a game. I have tested this using a gpu and cpu stress test. The cpu or gpu being stressed by themselves is fine. Stressing the cpu, and then starting the gpu stress test 10 seconds later, or vice versa, is fine. BUT when you start the gpu and cpu stress tests at THE EXACT SAME TIME.....BOOM computer restarts. It's 100% a power/voltage management issue.
 
Upvote 0 Downvote
Has there been any solution to this? Mine is for the most part, identical. Nothing out of the ordinary happens when I'm using the PC for basic applications, but when gaming, it's random. A few matches in nothing happens, but then i'll be playing something and screen will go black/pc will turn off (similar to holding the power on button till it shuts off), then it'll restart itself as if nothing happened. As others have said, no dump is created because its sudden. This issue never occurred until a few months back . After the comp starts up again, it runs as if nothing happened and I don't get any more shutdowns while gaming. It only happens generally once (maybe twice) on a single 'on' session (without manually shutting it down). The only thing I can think of doing is rolling back updates to just prior to spring update. Cause I'm not sure which update after it started all these problems. Any update would be helpful. I've also tried a PSU, GPU, and CPU stress test and alone, they all operate just fine.
 
Upvote 0 Downvote
Sorry for the late reply. The issue has persisted, UNTIL today, when it did one final crash. This time however, the computer did not automatically restart. I attempted to manually start the computer, to no avail. In the end, I THINK that it turned out to be my EVGA 980ti the whole time, and it had just been slowly failing. I have since removed the card, and as soon as I did the computer started up with no issues, with everything including ram still being properly detected.

It is curious though, the fans on the 980ti were working properly, it never went over 75c even on intense gaming, it was not overclocked, and it had been fine for over 2 years, it just started dieing, and is now dead, as far as I can tell. Luckily it is still in the 3 year warranty from EVGA, so I contacted them and am waiting on their response.

For now I am just using the integrated graphics in my 8700k cpu to run league, and the crash has not happened.....yet.

Unless the crash happens again after this point, or EVGA looks at the card and says its good, I am naming the 980ti the culprit as a dead piece of hardware.

I will be sure to ask for help again if the issue come back though, and again, thank you for all of you help!
 
Upvote 0 Downvote


Yea I can smell that typical burnt electronic smell coming from an area of its pcb that is near the eight and 6 pin power ports, which makes since if there was a capacitor or resistor or something that was going bad there. The earlier crashes/ restarts were probably the power supply detecting something strange going on so it would shut off the computer.
 
Upvote 0 Downvote

^Before you ruled out the 980ti, would it restart once then run normal again until you shut it down? (Gaming included)

 
Upvote 0 Downvote


Before I figured out it was the 980ti, it would do the crash then restart thing I mentioned earlier. Then yesterday it shutdown without restarting, and would not start until I physically removed the card. There were no indicators it was the card, temperatures and voltages were fine, driver and bios were up to date. Clearly the earlier crashes were the power supply' sensing that something was wrong and thus cutting power to protect the system. I suspect that it was a bad capacitor on the card or something, since the smell is coming from the area of the pcb that is next to the 8 and 6 pin power connectors.
 
Upvote 0 Downvote

Hrmmm I'll have a wiff at my board later then to smell if there is a burning odor. If so, it's off to RMA to MSI. Something to look at then. Thanks! My temps and such are also good oddly enough.

 
Upvote 0 Downvote
^So I smelled my card, and didn't pick up any burnt odors at all. I'm not sure how it worked for the other guy, but everytime it restarts, the PC works fine until the next time I shut it down and start it again. It seems to do that once each startup, and only during gaming.

UPDATE: So it was the PSU. I managed to get my hands on Corsair PSU, and it hasn't crashed once. I'm surprised that a slightly over than 1 year old EVGA Gold PSU would giving problems so soon. Least they'll honor an RMA.
 
Upvote 0 Downvote
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom