PC crashes ~30 minutes into SOME games, cooling issue?

[Infernalz]

For future generations looking for this:

The issue was drivers. We updated the BIOS, the chipset, some driver for the power mannagement, and the sound driver. Somewhere in there it was fixed, probably the power management and/or BIOS. We saw what drivers were missing/outdated somewhere in the system, he looked it up himself. We also changed out the thermal paste, the AIO had its little thin square pre-applied but when we took it off there was barely any left on the IHS to show where it made contact. After I put some of mine I had left it dropped his temps almost TWENTY DEGREES. Rediciouls.

 

This computer was bought by one friend maybe 2 years ago, a cyberpower pre-built, then is now with another friend, this issue was started around of it's last uses before the first friend upgraded recently and passed it on. The computer has a weird issue where it will BSOD and crash about 30 minutes into running only some games. Stellaris and The Forest die, while Warframe doesn't. Having him keep an eye on temps every 5 minutes while he plays shows that his temps rise over 80C on all cores at about that 30 minute mark. He is running an i9-9900K using some SLIM 1 x 120mm AIO cooler, I don't know the exact brand but it is sure a slim variant in that the whole radiator is as thin as the fan itself.

 

I can only assume this problem is in cooling, I can't imagine that slim cooler could handle that CPU and it taking a bit to heat up to fail is my only evidence. Are there any better ways to test this? Heck we might pull the AIO off the CPU and see the 'remove before installing' plastic bit and be done with it, but I'd like other opinions here. Any help would be great.

[514_AppleGuy]

Honestly, 80˚C is high, but I don't think it would be high enough to cause major throttling and crashes. You could try to improve cooling and see if it resolve the issue, but I really don't think it'll help. I would start by making sure that BIOS is the latest available version, and then I would run a stress-test on the RAM (maybe MemTest). 

 

From personal experience, all of the "crashes" I've ever experienced were either software related or due to RAM issues.

[Infernalz]

So I remembered the event viewer and got him to find what major went on when it crashed,

 

The computer has rebooted from a bugcheck.  The bugcheck was: 0x00000124 (0x0000000000000000, 0xffffde0996ca3028, 0x00000000b2000000, 0x0000000000030005). A dump was saved in: C:\Windows\MEMORY.DMP. Report Id: 6ebec81b-698d-423d-afa5-9891f9c6c694.

 

I couldn't find anything related to that bugcheck on my own.

[Infernalz]

Can a brother get a bump ;_; Need me some more ideas here.

[Infernalz]

Haha new info. Got a different error code involving 'kernel power,' of course an easy cooler change is never the issue.

 

- <System>
  <Provider Name="Microsoft-Windows-Kernel-Power" Guid="{331c3b3a-2005-44c2-ac5e-77220c37d6b4}" />
  <EventID>41</EventID>
  <Version>8</Version>
  <Level>1</Level>
  <Task>63</Task>
  <Opcode>0</Opcode>
  <Keywords>0x8000400000000002</Keywords>
  <TimeCreated SystemTime="2020-12-30T19:08:37.5997212Z" />
  <EventRecordID>7598</EventRecordID>
  <Correlation />
  <Execution ProcessID="4" ThreadID="8" />
  <Channel>System</Channel>
  <Computer>DESKTOP-3NBKPJU</Computer>
  <Security UserID="S-1-5-18" />
  </System>
- <EventData>
  <Data Name="BugcheckCode">127</Data>
  <Data Name="BugcheckParameter1">0x8</Data>
  <Data Name="BugcheckParameter2">0xffffd78024fe0e50</Data>
  <Data Name="BugcheckParameter3">0xffffd78024fdaef0</Data>
  <Data Name="BugcheckParameter4">0xfffff80049a03610</Data>
  <Data Name="SleepInProgress">0</Data>
  <Data Name="PowerButtonTimestamp">0</Data>
  <Data Name="BootAppStatus">0</Data>
  <Data Name="Checkpoint">0</Data>
  <Data Name="ConnectedStandbyInProgress">false</Data>
  <Data Name="SystemSleepTransitionsToOn">0</Data>
  <Data Name="CsEntryScenarioInstanceId">0</Data>
  <Data Name="BugcheckInfoFromEFI">false</Data>
  <Data Name="CheckpointStatus">0</Data>
  <Data Name="CsEntryScenarioInstanceIdV2">0</Data>
  <Data Name="LongPowerButtonPressDetected">false</Data>
  </EventData>
  </Event>

[Convictions]

Is the PSU sufficient?

[Infernalz]

21 minutes ago, Convictions said:

Is the PSU sufficient?

I can't imagine it's not given it was from a system builder, but the PSU sticker isn't facing the right way to know and he's not one to take it apart and see and I can't get there until tomorrow to try.

 

We're going to swap in his old PSU and see if that changes anything, newegg's power calculator shows only needing 462 watts for the system and his old one should be 550, technically above but not a lot of headroom.