Looking for advice about PC crashing

Phil87 · 21 Jul 2018 at 18:16

Hi all, so I've been having problems with my PC crashing recently, only in games so far. The two games I've had crashes from are Prey and Final Fantasy XV. PC specs are below in sig and I'm running Win10 Pro 1803, nothing is overclocked or overheating (CPU 65C or below in game, GPU a touch over 70C).

Prey crashed twice on the same day, both times it locked up for a while (5-30 seconds) and then the PC rebooted by itself (no blue screen). BlueScreenView showed one error related to the graphics driver and one related to ntoskrnl.exe i.e. a different error for each crash.

I just got FFXV this morning and it's crashed twice, once was a crash to desktop with an error in Event Viewer complaining about the game exe, and the other was a PC reset with BlueScreenView complaining about ntoskrnl.exe, and again there was no blue screen when the crash happened.

So I'm just after some advice about how to find what's causing the problem. I'll be running Memtest86 overnight (at least 8 passes) but recently replaced RAM (around 7 months ago) as one of the old sticks was giving errors. I did a quick run of Memtest86 with just 1 pass, no errors but got "Note - RAM may be vulnerable to high frequency row hammer bit flips" - am I right in thinking this is nothing to worry about?

I'm thinking the PSU is the most likely culprit, but before I replace that I was thinking of doing a reformat to rule out and software / driver problems. I'd be grateful for any advice.

jsmoke · 21 Jul 2018 at 19:45

Surely a driver or memory problem, maybe try with one stick of ram and vice versa etc.

On the driver side you can only update windows and try a different graphics driver.

Phil87 · 21 Jul 2018 at 20:00

I already updated the graphics driver since the crashes in Prey (that was a couple of weeks ago), played that for at least a few more hours since then with no issues but now of course I'm having a similar problem. I might try removing one stick but unfortunately FFXV uses over 10GB sometimes, so I'd have to lower the settings and see if the crash still occurs.

Are these the kind of problems that would also happen if the PSU is faulty, and is a crash where the PC reboots definitely not just a problem with the game itself?

jsmoke · 21 Jul 2018 at 20:46

I would have thought power or overheating issues would cause an instant reboot. Of course maybe there are exceptions. It just rings more with a memory or driver or other software conflict.

You sound as though you have all bases covered though. Is there nothing else you can learn from event viewer or the other app you mentioned such as error codes.

When you run out of options all you can do is start swapping components out such as GPU, ram and PSU etc.

Phil87 · 21 Jul 2018 at 21:01

Here's all the relevant error info I could find from Event Viewer and BlueScreenView, also posted this on the Steam forums and someone recommended disabling fast startup so that's the next thing I'm trying.

Error - EventLog
The previous system shutdown at 16:26:38 on ‎21/‎07/‎2018 was unexpected.

Error - Microsoft-Windows-EnhancedStorage-EhStorTcgDrv
A TCG Command has returned an error.
Desc: AuthenticateSession
Param1: 0x1
Param2: 0x60000001C
Param3: 0x900000006
Param4: 0x0
Status: 0x1

Critical - Microsoft-Windows-Kernel-Power
The system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly.

Error - Microsoft-Windows-WER-SystemErrorReporting
The computer has rebooted from a bugcheck. The bugcheck was: 0x00000154 (0xffffe18e1862a000, 0xffffa5031cea6840, 0x0000000000000001, 0x0000000000000000). A dump was saved in: C:\WINDOWS\MEMORY.DMP. Report Id: f8a652e0-95fe-4ae3-8bf4-40e3f9122bd7.

And some more info from BlueScreenView:

072118-6750-01.dmp
21/07/2018 4:27:22 PM
0x00000154
ffffe18e`1862a000
ffffa503`1cea6840
00000000`00000001
00000000`00000000
ntoskrnl.exe
ntoskrnl.exe+198430
x64
ntoskrnl.exe+198430
C:\WINDOWS\Minidump\072118-6750-01.dmp

Phil87 · 22 Jul 2018 at 08:12

So I ran Memtest86 V7.5 overnight, and when I turned on the monitor there was no signal although the PC was still doing stuff, I could turn num lock on and off and saw the light on the USB stick flashing a couple times. In the end I had to hold in the power button, then the PC started fine and I found this in the Memtest log:

Finished pass #10 (of 999) (Cumulative error count: 0)

So 10 passes with no errors, just that strange note about "high frequency row hammer bit flips". I guess that means the memory is ok and looking at the times in the log it was still running tests up to the point I turned off the PC, but the fact that there was no signal on the monitor concerns me. I did a bit of googling and some people are saying there should be no screen saver in Memtest or the BIOS, so could this indicate a problem?