Windows 10 / Computer Shuts Down Suddenly, OS Corrupted.

[RalphORama]

I recently built 2 identical computers for my workplace, and I've been having an issue with one of them.

When both sticks of RAM are inserted into the motherboard, the computer will, after a random amount of time, shut off suddenly and without warning. When I press the power button to turn it on again, it boots, but displays the error message "Your PC ran into an error and needs to restart," yadda yadda yadda. No matter how many times I let the OS restart, it displays this message.

I know the message gives the type of error, but the specific type has changed almost every time. Off the top of my head, the two errors I remember are system_pte_misuse and page fault in nonpaged area

I've temporarily solved the issue by plugging only one stick of RAM in. When I put it in the other slot, the computer did turn off unexpectedly once, but successfully booted again with no error.

My parts are:

• 
  - Intel i7-6700k
  - MSI Z170A Gaming 5
  - Corsair Vengeance LPX 16GB (2 x 8GB) DDR4-2400
  - Asus Radeon R9 Nano
  - Samsung 840 Evo (500GB)

Any idea how I could diagnose/determine the source of this issue?

 

[Colif]

run http://www.memtest86.com/ on the ram
update firmware on the evo using samsung magician
update bios on motherboard
Update drivers as system_pte_misuse can be drivers

follow the instructions on this page (http://www.tenforums.com/tutorials/5560-bsod-minidump-configure-create-windows-10-a.html)
Do option one: Configure Dump File Settings in Start-up and Recovery
and then do Have Windows Create a Small Memory Dump (Minidump) on BSOD in Small minidumps

this will create a file in c/windows/minidump
copy the file to your documents folder and then upload that copy to a cloud server and share link here (you cannot upload direct from minidump folder). Someone will read the logs and figure out problem

download and run http://www.resplendence.com/whocrashed
it will look at error logs and create a summary

please copy/paste the summary into this thread.

 

[RalphORama]

<snip>

Thank you for your reply! I'll give all this a shot when I get into the office tomorrow.

 

[RalphORama]

OK, so I've done some testing...
I installed only one stick of RAM, and that seemed to mitigate the issue long enough to allow me to install the proper drivers and get Windows 10 running. I then did the following:
- Installed all the proper driver software
- Checked for an update to the Evo firmware (there wasn't one, it's up to date)
- Downloaded and installed BIOS update version F4

After all this, I ran a 4-pass memtest, and it didn't find any errors. However, when set to use any CPU besides "Single: CPU #0," memtest crashes just after starting the tests. I've tried the following:
- Parallel (All CPUs)
- Round Robin (All CPUs)
- Sequential (All CPUs)
- Single: CPU #1 through #7
Each of these settings caused memtest to crash and the system to reboot.

There is no minidump file (as Windows hasn't crashed yet), but I'm working on generating one. Also, WhoCrashed didn't output anything useful, as there was no minidump file.

 

[Colif]

it appears the memtest itself is not stable on multi cores. I didn't know that detail.

who crashed looks at error logs, not the minidump. your log files can be massive, its why we only ask for the minidump as its only 256kb, the logs can be a few gb in size.

 

[RalphORama]

Well, the issue is that Windows doesn't BSOD, the computer itself shuts off abruptly. Therefore, WhoCrashed is not able to tell me what's going wrong, as no crash dump is generated.

The one way I've been able to reproduce this reliably is to download a large file using Google Chrome. If I boot up the computer and set Google Chrome to download the ISO of Visual Studio Community 2015, after about 30-60 seconds, the computer will crash. I'll see if I can get a video of this behavior and post it in this thread.

 

[Colif]

So all sticks work but not in the pc at same time. did you check different slots as it could be the motherboard.
could be a power thing too, not enough power for pc to run the ram, I don't think its the psu, more just a setting. I don't know that much about ram voltages to make a decision really.

guess quicker test of ram would have been to just see if it worked in other pc, i should have thought of that before.

 

[RalphORama]

I'll give Gigabyte a call and see if there's anything in their knowledgebase about this issue. Thank you very much for your help!

Side note: I ran memtest on the PC that works and experienced the same crashes when trying to use alternate cores. I think it's a memtest issue and not a hardware issue.