Windows 10 randomly restarts (no BSOD) - possibly while gaming

[tgrmst]

My PC has been restarting randomly lately... mostly during gaming load (90+% load, 60+C temps). No BSOD

Not sure where to start, or even if this is the right section... BIOS updated. Drivers updated. Here is spec:

Windows 10 
6700k
Gigabyte Z170 GT
1080 Ti FE
G Skills Ripjaw 3200 - 32GB
Seasonic 860W Platinum

Thoughts?

 

[Danny_UK]

Bad overclock? (If overclocked things can degrade and go bad. Though this rare if less than 6 months)

[tgrmst]

Just now, Danny_UK said:

Bad overclock? (If overclocked things can degrade and go bad. Though this rare if less than 6 months)

I reset everything to default. No overclocks at all. Even RAM is at default 2133 MHz, not the XMP profile.

[Danny_UK]

4 minutes ago, tgrmst said:

I reset everything to default. No overclocks at all. Even RAM is at default 2133 MHz, not the XMP profile.

Does it happen on specific titles? How long on avg before the game will crash/pc restarts? Have you tried running Aida64 heat/stress (FPU) test to see if the pc reboots then? 

 

Is this a torrented game? >_> (I'm not judging)

[tgrmst]

4 minutes ago, Danny_UK said:

Does it happen on specific titles? How long on avg before the game will crash/pc restarts? Have you tried running Aida64 heat/stress (FPU) test to see if the pc reboots then? 

 

Is this a torrented game? >_> (I'm not judging)

Actually now it is not gaming related. But even on those games, it was random, and on both legit steam and torrented ones (no offense taken). 

 

I did Heaven Benchmark and Furmark before, no issues after 30 mins. Also Intel Extreme Tuning Utility to stress test CPU/RAM. Again 30 mins no issue.

 

Funny thing is it restarted just now while I was trying to reply to you. No game, while browsing. Youtube was playing for a while, but stopped then it restarted...

[Danny_UK]

Humm, how old is your O/S disk? Is your PSu warm?

 

Tbh i was gonna say upload a crash dump and i'll take a quick look at it for you, but... I work nights and I'm struggling to stay awake atm, I'm usually in bed like 20mins ago.

 

I was hoping it was gonna be something simple and someone else would have jumped in and helped you troubleshoot.  I mean those are the common ones that are easy to detect, other common issues are driver/windows updates.  

 

If you want you can upload your crashdump (.DMP) and when i wake back up in 6hours or so ill take a look at it for you, if no one else has by that time, or managed to find out why.

 

Sorry to be like, "Yo whats up? Oh ok, im going now bye"  

I'll get back to you if you don't get a solution by the time i've woke up.

[Ansuex]

Hii,

 

I have kind of had the same issue.

My old i7 920 system seems to work flawless, but then it crashes out of random sometimes.

Now after a longgg time of testing I found the cause. My cooler backplate was shorting out the motherboard resulting in random crashes and reboots.

 

Maybe this might be the same for you?

It could as well be that some power cables are not seated correctly or maybe you have faulty hardware?

I think you should atleast go with a test bench outside of your case and see what happens.

[tgrmst]

1 minute ago, Danny_UK said:

Humm, how old is your O/S disk? Is your PSu warm?

 

Tbh i was gonna say upload a crash dump and i'll take a quick look at it for you, but... I work nights and I'm struggling to stay awake atm, I'm usually in bed like 20mins ago.

 

I was hoping it was gonna be something simple and someone else would have jumped in and helped you troubleshoot.  I mean those are the common ones that are easy to detect, other common issues are driver/windows updates.  

 

If you want you can upload your crashdump (.DMP) and when i wake back up in 6hours or so ill take a look at it for you, if no one else has by that time, or managed to find out why.

 

Sorry to be like, "Yo whats up? Oh ok, im going now bye"  

I'll get back to you if you don't get a solution by the time i've woke up.

LOL perfectly ok good sir. I should be in bed too (1AM for me). 

 

OS disk is pretty new... about a year. The whole system was a slow upgrade part by part about 2 years ago. ADATA 120GB. 

I will need to check temps and get you the dump later. 

Just now, Ansuex said:

Hii,

 

I have kind of had the same issue.

My old i7 920 system seems to work flawless, but then it crashes out of random sometimes.

Now after a longgg time of testing I found the cause. My cooler backplate was shorting out the motherboard resulting in random crashes and reboots.

 

Maybe this might be the same for you?

It could as well be that some power cables are not seated correctly or maybe you have faulty hardware?

I think you should atleast go with a test bench outside of your case and see what happens.

I am guessing it is a hardware issue... but it was really weird since I didn't do any recent upgrade... but something might be intermittent for a while now... I will open and check it tomorrow morning. Will report back with results. 

 

Thanks guys. If anyone else have ideas, please leave them. I will continue troubleshooting in the morning. Thanks again.

[tgrmst]

AM update:

 

I won't be posting the dump file. It was last updated in April, and file size is 1.4 gigs. I didn't get a BSOD, so I'm pretty sure it was not logged as all, since it was not a safe crash that Windows initiated to recover from, but more of a hardware issue...

 

I ended up opening the case, unplug/plug in all the power cables on the drives and the PSU. Re-seat the RAMs and GPU. Also redid the thermal compound that I should have done last time. Pretty much trying to get all the contacts to be fresh again. Dusted and put it all back. Ran a game for about 10 mins. Nothing weird. No reboot after 1 hour of regular use yet... Going to let it be for about a week, and but I'll post again if it does reboot. 

[johndms]

24 minutes ago, tgrmst said:

AM update:

 

I won't be posting the dump file. It was last updated in April, and file size is 1.4 gigs. I didn't get a BSOD, so I'm pretty sure it was not logged as all, since it was not a safe crash that Windows initiated to recover from, but more of a hardware issue...

 

I ended up opening the case, unplug/plug in all the power cables on the drives and the PSU. Re-seat the RAMs and GPU. Also redid the thermal compound that I should have done last time. Pretty much trying to get all the contacts to be fresh again. Dusted and put it all back. Ran a game for about 10 mins. Nothing weird. No reboot after 1 hour of regular use yet... Going to let it be for about a week, and but I'll post again if it does reboot. 

You could always attempt to 'force' the problem to happen when you're watching. Often when people complain of random, unexplained crashes, I recommend running several separate stress tests using Aida64's System Stability Test. My usual technique is to start with the CPU. Select the first three boxes (CPU / FPU / Cache) and run the test for at least 1 hour. If it doesn't fail, stop the test and uncheck those three boxes. You can then check the System Memory box and stress that for one hour.

 

Note: I've begun using Memtest86 instead of Aida64 to test ram. The method I use is 10 runs of test #6 with 'all cores parallel'. Just look around Memtest86's options and you'll see how to disable all tests but #6. This usually takes me about 20 minutes. If it passes, then I'm comfortable saying the ram isn't the problem. This method comes from Comprehensive Memory Overclocking Guide, and also includes a Prime95 test if you want to make be absolutely positive ram isn't the problem.

 

Next, you can use Aida64 to stress the GPU for an hour. Then you can test all of the components at the same time for an hour. If this method is unable to force a problem, then maybe check the SMART info on the drive, or maybe try another PSU. It'd be strange to have a SeaSonic Platinum become faulty, but who knows? Hmm, may be boot into Safe Mode and see if any issues happen there.

[Danny_UK]

1 hour ago, tgrmst said:

AM update:

 

I won't be posting the dump file. It was last updated in April, and file size is 1.4 gigs. I didn't get a BSOD, so I'm pretty sure it was not logged as all, since it was not a safe crash that Windows initiated to recover from, but more of a hardware issue...

 

I ended up opening the case, unplug/plug in all the power cables on the drives and the PSU. Re-seat the RAMs and GPU. Also redid the thermal compound that I should have done last time. Pretty much trying to get all the contacts to be fresh again. Dusted and put it all back. Ran a game for about 10 mins. Nothing weird. No reboot after 1 hour of regular use yet... Going to let it be for about a week, and but I'll post again if it does reboot. 

If you got it to work bro that's all that counts. Also Win10 doesn't always "Blue Screen", I've only encountered that once trying to run my memory at 3400mhz, but when windows does halt and restarts no doubt it's written something to the dump file.

 

Anyhow, Good Job with the fix

[tgrmst]

7 hours ago, johndms said:

You could always attempt to 'force' the problem to happen when you're watching. Often when people complain of random, unexplained crashes, I recommend running several separate stress tests using Aida64's System Stability Test. My usual technique is to start with the CPU. Select the first three boxes (CPU / FPU / Cache) and run the test for at least 1 hour. If it doesn't fail, stop the test and uncheck those three boxes. You can then check the System Memory box and stress that for one hour.

 

Note: I've begun using Memtest86 instead of Aida64 to test ram. The method I use is 10 runs of test #6 with 'all cores parallel'. Just look around Memtest86's options and you'll see how to disable all tests but #6. This usually takes me about 20 minutes. If it passes, then I'm comfortable saying the ram isn't the problem. This method comes from Comprehensive Memory Overclocking Guide, and also includes a Prime95 test if you want to make be absolutely positive ram isn't the problem.

 

Next, you can use Aida64 to stress the GPU for an hour. Then you can test all of the components at the same time for an hour. If this method is unable to force a problem, then maybe check the SMART info on the drive, or maybe try another PSU. It'd be strange to have a SeaSonic Platinum become faulty, but who knows? Hmm, may be boot into Safe Mode and see if any issues happen there.

I don't really have chance to get it to run as intense and long as you wanted it. I cannot leave it unattended for too long, so I did 30 min for the first test, and 30 for the memory test in Aida64. No issues. No reboot.

 

Also I just got back and did some gaming for about 30 mins, the same game that crashed and rebooted before. No issues again. 

 

I will stress it out again should it fail, but for now I think it is gone... 

6 hours ago, Danny_UK said:

If you got it to work bro that's all that counts. Also Win10 doesn't always "Blue Screen", I've only encountered that once trying to run my memory at 3400mhz, but when windows does halt and restarts no doubt it's written something to the dump file.

 

Anyhow, Good Job with the fix

Thanks and I agree that there would be something written to the dump file even without BSOD, but it definitely didn't act like it so I doubt there is anything in there.

 

BTW, one thing I remember while debuilding was that the CPU fan on my Noctua NH-D15 was sitting directly on the RAM. One of my thoughts is that the fan vibration may have messed up with the RAM, so I moved it higher...

 

So far no issues yet. I can mess with it for a few more hours. Hope it is fixed.

[tgrmst]

Over a week update:

 

No issues since. I'm guessing whatever I did with the hardware reseating/rebuild/remount fixed it. I have not try to OC or do anything but game/browse and some light video processing. Thanks to all for your help.

[PhaseShift]

Start removing peripherals one by one to isolate the issue (if there is still one). My friend's pc was randomly shutting down and after 2 months it turned out his crappy mouse was sending some kind of a bad signal to the motherboard and that initiated shutdown . So yeah hope this helps!