PC Reboots (no BSOD) when playing fairly intensive games

[Level 3 Mage]

Hi Everyone,

I posted to Toms hardware but no response, I am hoping someone can help. 

 

OS: Windows 10 Pro x64
Motherboard: ASUS Sabertooth 990FX R2.0
CPU: AMD FX 6300 Bulldozer
RAM: 16GB (2 x 4GB Corsair Vengeance, 2 x 4GB Patriot)
Graphics Card: XFX RX480 8GB
PSU: Corsair CX750
HDD's:-
1. 128GB Samsung SSD (OS)
2. 500 GB HDD (Files and Documents)
3. 2TB HDD (Games)
4. 128GB Samsung SSD (Offline Games)

I have recently started facing an issue where my PC will reboot whilst playing fairly graphic intensive games (Destiny 2, For The King, Rainbow Six Siege), checking the event logs I just get a Critical Kernel 41 error. 
I am not over heating, when I reboot my CPU it is at 55c and Graphics is about 60c (which is pretty normal).

I have reinstalled all graphic drivers, CPU drivers, Windows Updates to no avail.
Windows reset and a format of my C drive with a fresh Windows 10 Pro image on it has not resolved my issue.

I have run HWMONITOR & recording whilst playing Destiny 2 for an hour and it was fine but once I was not recording it rebooted in about 10 minutes. Whilst watching HWMONITOR I do not see any issues with the PSU fluctuating high or low, my temps are always at a good margin.

I have run some stress test software (CPU RAM and GPU) which has not crashed running for over an hour

I do notice however that when it does reboot my keyboard and mouse (Steelseries RGB) will be off and then flicker hard, sometimes I am unable to get into the BIOS as they keyboard and mouse will not come on until I am in the windows loading screen.

Any help or suggestion would be appreciated, 
Thanks

p.s. I have given the machine a deep clean. Games like CSGO are fine and I have no issue with.


************************************************************
UPDATE
************************************************************
Running PRIME95 and Furmark at the same time for about 5 minute crashed my PC.
However running both individually will not cause a crash.
I have just tested with another PSU and I am still having the same problem running both PRIME95 and Furmark.
Tested a new cooler and CPU and now whenever I run Furmark it will instantly reboot. 
I have also downgraded my GPU drivers and the same issue.
Tried a new MB and still the same issue. Only thing left is my GPU

[KuJoe]

If it's only crashing (not BSOD) when you use your videocard then it sounds like it's power related. You said you tried another PSU already, what PSU did you try? 750W should be enough but your PSU might be going bad. Have you tried plugging into a UPS to "clean" the power before it goes to your PC?

 

EDIT: If you don't already have one, I'd go pick up a Kill-a-watt meter off Amazon for $15-$20. They are extremely useful and can help determine whether you're having power issues with your PC.

[Morgan MLGman]

Yeah, this is most likely the culprit:

Quote

Corsair CX750

What was the other PSU that you tested with?

[Level 3 Mage]

Hi KuJoe,

 

Thanks for getting back to me. 

 

I have tried another PSU (was a Zalman own brand but known to be perfectly fine and working) and there was no change when running FurMark (reboots in the space of 10 minutes)

I have noticed when it reboots sometimes the mouse and keyboard wont light up sometimes until windows is half way booted.

I wont be able to test with a UPS unfortunately. 

 

What readings should I be seeing with the Kill-a-watt meter?

 

Thanks,

 

[homeap5]

Install MSI Afterburner and underclock your GPU (-400 - both clock and memory). Then try.

 

41 minutes ago, Level 3 Mage said:

Windows reset and a format of my C drive with a fresh Windows 10 Pro image on it has not resolved my issue.

Why the hell everyone destroy everything instead of looking for a problem? In every thread when someone has a problem, I read "I made format / refresh / reinstall / clean install". How this may help? Second time Windows write different files to hard drive? Don't do that again, find problem instead of blind shooting. Or at least ask first, then made reinstall if everything else fails.

 

[Level 3 Mage]

18 minutes ago, Morgan MLGman said:

What was the other PSU that you tested with?

It was a Zalman and a XFX 450w and this made no change

[Level 3 Mage]

6 minutes ago, homeap5 said:

Install MSI Afterburner and underclock your GPU (-400 - both clock and memory). Then try.

 

Why the hell everyone destroy everything instead of looking for a problem?

I will install this and try those settings. If this is stable I assume it would be my PSU?

 

I needed a fresh install anyway, I make it a yearly thing. I only use a 128GB SSD for my OS, all my other files are on the other drives.

[homeap5]

47 minutes ago, Level 3 Mage said:

I will install this and try those settings. If this is stable I assume it would be my PSU?

 

I needed a fresh install anyway, I make it a yearly thing. I only use a 128GB SSD for my OS, all my other files are on the other drives.

It's probably your GPU, not PSU.

And don't make it a "yearly thing". It's kind of strange and not necessary at all. Unless you doing nothing interesting on your computer and it's just your hobby. Sorry, but that myth (that Windows need to be reinstalled frequently) is annoying (Mac and Linux users have fun).

[Level 3 Mage]

6 minutes ago, homeap5 said:

It's probably your GPU, not PSU.

And don't make it a "yearly thing". It's kind of strange and not necessary at all. Unless you doing nothing interesting on your computer and it's just your hobby. Sorry, but that myth (that Windows need to be reinstalled frequently) is annoying (Mac and Linux users have fun).

Thanks Ill try to do this less often. 

I'm installing MSI afterburner now and will try those settings out.

 

Thanks a lot for your help guys.

[KuJoe]

16 minutes ago, homeap5 said:

It's probably your GPU, not PSU.

If it's the videocard it would cause the system to crash, not reboot. There are very few things that would cause a PC to power off without an error, most of them are power related.

[Level 3 Mage]

5 minutes ago, KuJoe said:

If it's the videocard it would cause the system to crash, not reboot. There are very few things that would cause a PC to power off without an error, most of them are power related.

Could the GPU be pulling too much power and causing an overload? 

Im not seeing any fluctuations in HWMONTIOR

[homeap5]

Please, try that MSI Afterburner first, then we can talk about theory that GPU cannot made system reboot or not. In my opinion and experience - it can.

[Level 3 Mage]

26 minutes ago, homeap5 said:

Please, try that MSI Afterburner first, then we can talk about theory that GPU cannot made system reboot or not. In my opinion and experience - it can.

So all installed and monitoring the clock speeds and temps. I am noticing my clock speeds on idle are all over the place. 300-1200Mhz every second at-least. When i lower the voltage it stabilises for me to around 800-900Mhz. 

Just tried running Furmark and it has instantly rebooted when i click run.

I can only lower the voltage It refuses me to change the MHZ for clock and RAM speeds

[KuJoe]

1 hour ago, Level 3 Mage said:

Could the GPU be pulling too much power and causing an overload? 

Im not seeing any fluctuations in HWMONTIOR

That's usually what it is. Your videocard uses the most power out of all of the components in your system, it also pulls the most power when you're gaming which is why you're only seeing the reboots then and are able to run for hours without a reboot when you're not gaming (some benchmarks will pull a lot of power also).

 

1 hour ago, homeap5 said:

In my opinion and experience - it can.

It can if it's power related. If the videocard is broken and causing the system to short circuit (i.e. power related), then it can cause a reboot. If it's not power related the system will not reboot though, it will either BSOD with an error or freeze the image on the screen while the other components continue running.

[homeap5]

2 hours ago, Level 3 Mage said:

So all installed and monitoring the clock speeds and temps. I am noticing my clock speeds on idle are all over the place. 300-1200Mhz every second at-least. When i lower the voltage it stabilises for me to around 800-900Mhz. 

Just tried running Furmark and it has instantly rebooted when i click run.

I can only lower the voltage It refuses me to change the MHZ for clock and RAM speeds

Why you lowering voltage?

Lower only clock and ram speed of your GPU, not voltage!

EDIT: sorry, now I see that you cannot change GPU MHz or GPU RAM speed. Well, that is bad. Maybe other, dedicated tool for your graphics card, can do that. You can also increase power limit (from 100 to the max - 114% probably) and see what happens.

PS. Try to force vertical sync in radeon control panel. It should decrease GPU usage and increase stability.

 

1 hour ago, KuJoe said:

It can if it's power related. If the videocard is broken and causing the system to short circuit (i.e. power related), then it can cause a reboot. If it's not power related the system will not reboot though, it will either BSOD with an error or freeze the image on the screen while the other components continue running.

I know what I saw few times, so in theory is not possible, but it happens anyway. Bad drivers or faulty GPU can made system reset, not only BSOD, even if system should handle it.

[Level 3 Mage]

2 hours ago, homeap5 said:

Why you lowering voltage?

Lower only clock and ram speed of your GPU, not voltage!

EDIT: sorry, now I see that you cannot change GPU MHz or GPU RAM speed. Well, that is bad. Maybe other, dedicated tool for your graphics card, can do that. You can also increase power limit (from 100 to the max - 114% probably) and see what happens.

PS. Try to force vertical sync in radeon control panel. It should decrease GPU usage and increase stability.

 

I know what I saw few times, so in theory is not possible, but it happens anyway. Bad drivers or faulty GPU can made system reset, not only BSOD, even if system should handle it.

Hi Homeap5,

 

Thanks for all your help on this I will try this once I get back from work. Should be home in an hour or so. Ill post back if I am still having issue. 

 

Thanks,

 

[Level 3 Mage]

2 hours ago, Level 3 Mage said:

Hi Homeap5,

 

Thanks for all your help on this I will try this once I get back from work. Should be home in an hour or so. Ill post back if I am still having issue. 

 

Thanks,

 

Just swapping back to my original motherboard now as I can pinpoint the issue which will require fresh install of Windows 10. 

I will get a base point setup and do some testing. I was able to increase the frequency and memory and have a stable Furmark running to which it would eventually randomly reboot. 

 

I have also logged a ticket with AMD to see if they can take some diagnostics from it and assist.

[homeap5]

10 hours ago, Level 3 Mage said:

Just swapping back to my original motherboard now as I can pinpoint the issue which will require fresh install of Windows 10. 

I will get a base point setup and do some testing. I was able to increase the frequency and memory and have a stable Furmark running to which it would eventually randomly reboot. 

 

I have also logged a ticket with AMD to see if they can take some diagnostics from it and assist. 

So, where is the problem? You said that you can pinpoint the issue.

And why another fresh install? Because motherboard swap? It's also not needed, unless you KNOW that it's system fault somehow.

 

[Astarore]

One thing that does strike me as odd is the fact that you could not change the clock speed of the GPU and only the voltage.I have seen a computer that would shut off randomly under load that was caused by a short in the GPU. Some power supplies do have over voltage protection that will shut off the PSU if a component is drawing way too much current. Perhaps that is what is happening here considering that you have tried other known components and the issue remained. 

[homeap5]

28 minutes ago, Astarore said:

One thing that does strike me as odd is the fact that you could not change the clock speed of the GPU and only the voltage.I have seen a computer that would shut off randomly under load that was caused by a short in the GPU. Some power supplies do have over voltage protection that will shut off the PSU if a component is drawing way too much current. Perhaps that is what is happening here considering that you have tried other known components and the issue remained. 

This is nothing unusual - MSI Afterburner sometimes don't support some options depends on card. That is why I asked for other dedicated tool for this card (I don't know is available).

[ewitte]

I would upload your crashdump and see if you get more answers

http://www.osronline.com/page.cfm?name=analyze

 

I've had it TWICE come back that BSOD was related to afterburner.  Seems they get a stable build then after a certain amount of time something with windows updates breaks it and I have to stop using it for a few months.

[homeap5]

2 hours ago, ewitte said:

I would upload your crashdump and see if you get more answers

http://www.osronline.com/page.cfm?name=analyze

 

I've had it TWICE come back that BSOD was related to afterburner.  Seems they get a stable build then after a certain amount of time something with windows updates breaks it and I have to stop using it for a few months.

He install Afterburner as my advice to underclock GPU, so this program is not a problem in this case.

[ewitte]

16 hours ago, homeap5 said:

He install Afterburner as my advice to underclock GPU, so this program is not a problem in this case.

Its still a good way to analyze a dump file it usually points to which DLL caused it.