PC restarting when playing games

[WellThen]

So, for a while now my pc has been crashing whenever playing games such as apex legends and quake champions, seems fine when I play stuff like planescape torment, though it is not limited to how intensive a game is, i think, since I still get crashes when playing a game which is not as intensive, such as reflex arena. When the crash happens, I get a black screen for a bit, then it just restarts. There are no error messages

 

System specs are as follows:

Windows 10 64bit

PowerColor Radeon RX 5700 XT 8 GB Red Devil Video Card

AMD Ryzen 5 3600 3.6 GHz 6-Core Processor

MSI B450 TOMAHAWK MAX ATX AM4 Motherboard   Corsair Vengeance LPX 16 GB (2 x 8 GB) DDR4-3200 CL16 Memory

Pioneer 1 TB M.2-2280 NVME Solid State Drive

 

Corsair RM650x 650 watts

 

Bios version is; BIOS Version/Date    American Megatrends Inc. 3.50, 07/11/2019

 

I tried swapping out a PSU with another Corsair TM-M 650 watt I had lying in the back and it did not help. Clean installing GPU drivers did not help. Run MemTest86, and it found no errors. I also error checked my ssd and hard drive, they seemed fine. Also tried seeing the temps at the time of crash, and the highest they went to was 87oC, but were at around 85 mostly. 

 

 

 

[Bombastinator]

33 minutes ago, WellThen said:

So, for a while now my pc has been crashing whenever playing games such as apex legends and quake champions, seems fine when I play stuff like planescape torment, though it is not limited to how intensive a game is, i think, since I still get crashes when playing a game which is not as intensive, such as reflex arena. When the crash happens, I get a black screen for a bit, then it just restarts. There are no error messages

 

System specs are as follows:

Windows 10 64bit

PowerColor Radeon RX 5700 XT 8 GB Red Devil Video Card

AMD Ryzen 5 3600 3.6 GHz 6-Core Processor

MSI B450 TOMAHAWK MAX ATX AM4 Motherboard   Corsair Vengeance LPX 16 GB (2 x 8 GB) DDR4-3200 CL16 Memory

Pioneer 1 TB M.2-2280 NVME Solid State Drive

 

Corsair RM650x 650 watts

 

Bios version is; BIOS Version/Date    American Megatrends Inc. 3.50, 07/11/2019

 

I tried swapping out a PSU with another Corsair TM-M 650 watt I had lying in the back and it did not help. Clean installing GPU drivers did not help. Run MemTest86, and it found no errors. I also error checked my ssd and hard drive, they seemed fine. Also tried seeing the temps at the time of crash, and the highest they went to was 87oC, but were at around 85 mostly. 

 

 

 

Most common hardware causes for crash after boot (which does not in any way imply the only causes) are memory and PSU doing a couple full runs of memtest86 can do a decent job of checking memory. It’s cheap but not fast. I don’t have any good solutions for PSU checking. 
 

If the memory come up clean from the test and you go on but still can’t solve things there is a more sensitive test but it take a lot longer.  You put a single stick of memory in slot one, run the thing, and see if it crashes again.

[WellThen]

On 6/10/2021 at 5:14 PM, Bombastinator said:

Most common hardware causes for crash after boot (which does not in any way imply the only causes) are memory and PSU doing a couple full runs of memtest86 can do a decent job of checking memory. It’s cheap but not fast. I don’t have any good solutions for PSU checking. 
 

If the memory come up clean from the test and you go on but still can’t solve things there is a more sensitive test but it take a lot longer.  You put a single stick of memory in slot one, run the thing, and see if it crashes again.

Which slot would be slot 1?  the manual just says DIMMB2, DIMMB1, DIMMA2, DIMMA1 and state s to put the stick into DIMMA2 if using single one. 

[Bombastinator]

1 hour ago, WellThen said:

Which slot would be slot 1?  the manual just says DIMMB2, DIMMB1, DIMMA2, DIMMA1 and state s to put the stick into DIMMA2 if using single one. 

Then you would use dimma2 I guess.  The use of a single stick ensures that most of the memory gets used while also eliminating confusion about which stick has the problem if any.  Simply running a game should use all the available memory so if there is a memory problem it should show up pretty fast.

Edited June 14, 2021 by Bombastinator

[WellThen]

16 hours ago, Bombastinator said:

Then you would use dimma2 I guess.  The use of a single stick ensures that most of the memory gets used while also eliminating confusion about which stick has the problem if any.  Simply running a game should use all the available memory so if there is a memory problem it should show up pretty fast.

So, one of the RAM sticks crashed the pc, the other did not.

[Bombastinator]

1 hour ago, WellThen said:

So, one of the RAM sticks crashed the pc, the other did not.

So you’ve possibly got a bad ram stick.  I might want to retest the good one to make really sure they’re not both crashing (in which case it’s not the dimms) but if it is replacing at least that ram stick (might work better replacing both because of paired ram and how things are packaged, etc..) may solve your issue

[WellThen]

5 hours ago, Bombastinator said:

So you’ve possibly got a bad ram stick.  I might want to retest the good one to make really sure they’re not both crashing (in which case it’s not the dimms) but if it is replacing at least that ram stick (might work better replacing both because of paired ram and how things are packaged, etc..) may solve your issue

oh well, when trying to run apex legends on the stick that seems to have been working it crashed . 

[Bombastinator]

19 minutes ago, WellThen said:

oh well, when trying to run apex legends on the stick that seems to have been working it crashed . 

So likely not a memory problem unless apex always crashes on 8gb machines.  I suspect it doesn’t.

[Demonic Donut]

Are you OCing any components? Even XMP?

[Bombastinator]

33 minutes ago, Demonic Donut said:

Are you OCing any components? Even XMP?

Good question.  I assumed no but an unstable OC could explain everything.

[Demonic Donut]

4 hours ago, Bombastinator said:

Good question.  I assumed no but an unstable OC could explain everything.

Yea I was curious because my ram ran fine at XMP for 2 years but started causing occasional crashing until I customized my OC for it.

[WellThen]

14 hours ago, Demonic Donut said:

Are you OCing any components? Even XMP?

The only thing I've changed when I built it was putting the XMP profile to 3200MHz,  though I went into bios now to double check and it was off, my memory being clocked at 2133MHz.

 

[Demonic Donut]

4 hours ago, WellThen said:

The only thing I've changed when I built it was putting the XMP profile to 3200MHz,  though I went into bios now to double check and it was off, my memory being clocked at 2133MHz.

 

This usually happens when you have an unstable OC and the bios will reset to try to correct it.

 

Could be bad RAM or even a memory controller issue. IMO bad RAM is more likely. It's the only component I've had somewhat regular issues with.

 

I would try resetting the bios manually to be sure to clear out anything strange, then up your RAM voltage to see if it can stabilize. 1.2V is standard, most XMP profiles run at 1.35V. The stock answer is don't go over 1.4V but memory OCers will go up to 1.5V assuming good case ventilation for cooling. I do try to stick under 1.4V personally.

 

You tried both sticks of RAM by themselves and they both crashed, yes? Both sticks *could* be bad. Depending on budget, you could buy a new set of sticks that are ryzen compatible and see if they work ok.

 

Also, have you ran memtest for a very long run? Try different settings and let it run overnight at least for a few nights to see if you can get an error to occur. Prime95 also has some good tests that will stress RAM and the memory controller.

 

Do you have HWInfo64 for monitoring? It will give you a lot more information than basic CPU temps.

[WellThen]

3 hours ago, Demonic Donut said:

This usually happens when you have an unstable OC and the bios will reset to try to correct it.

 

Could be bad RAM or even a memory controller issue. IMO bad RAM is more likely. It's the only component I've had somewhat regular issues with.

 

I would try resetting the bios manually to be sure to clear out anything strange, then up your RAM voltage to see if it can stabilize. 1.2V is standard, most XMP profiles run at 1.35V. The stock answer is don't go over 1.4V but memory OCers will go up to 1.5V assuming good case ventilation for cooling. I do try to stick under 1.4V personally.

 

You tried both sticks of RAM by themselves and they both crashed, yes? Both sticks *could* be bad. Depending on budget, you could buy a new set of sticks that are ryzen compatible and see if they work ok.

 

Also, have you ran memtest for a very long run? Try different settings and let it run overnight at least for a few nights to see if you can get an error to occur. Prime95 also has some good tests that will stress RAM and the memory controller.

 

Do you have HWInfo64 for monitoring? It will give you a lot more information than basic CPU temps.

Ok, will do the things suggested. What else, other then the temps, would I be looking for in HWInfo64? Got it to crash with it running and the log said it run at 73oC

[WellThen]

On 6/16/2021 at 3:53 PM, Demonic Donut said:

This usually happens when you have an unstable OC and the bios will reset to try to correct it.

 

Could be bad RAM or even a memory controller issue. IMO bad RAM is more likely. It's the only component I've had somewhat regular issues with.

 

I would try resetting the bios manually to be sure to clear out anything strange, then up your RAM voltage to see if it can stabilize. 1.2V is standard, most XMP profiles run at 1.35V. The stock answer is don't go over 1.4V but memory OCers will go up to 1.5V assuming good case ventilation for cooling. I do try to stick under 1.4V personally.

 

You tried both sticks of RAM by themselves and they both crashed, yes? Both sticks *could* be bad. Depending on budget, you could buy a new set of sticks that are ryzen compatible and see if they work ok.

 

Also, have you ran memtest for a very long run? Try different settings and let it run overnight at least for a few nights to see if you can get an error to occur. Prime95 also has some good tests that will stress RAM and the memory controller.

 

Do you have HWInfo64 for monitoring? It will give you a lot more information than basic CPU temps.

Ok, so I did all of the above, except of getting new ram. I just got some 8gigs of ram from my sister's pc and changed it with mine, still crashed. 

Resetting bios did not work, even when running at stock ram profile.  

Mem test came up with nothing. 

Did 4 runs of memtest86 with different settings and no errors.