Memory and other hardware issues...

Windows 10 Pro, 64-bit, version 2004 (OS Build 19041.630)...

I haven't been dealing with hardware issues for long time, kind of rusty, could use some help...

Couple of weeks ego, the system started to throw BSODs and the issue was related to "MEMORY_MANEGEMENT". A quick run of the MemTest86 showed, that the test failed. Here are the results:



In total, there are 63 errors. Getting new memory to address this issue. Interestingly, during this week, the BSODs did not return.

On the other hand, different problems started to show. For example, the system would have the Outlook ('13) open, writing an email locks up Outlook, then the system goes down hard. There's no BSOD, the system just stops, like when you loose the power and you have no UPS. System booted up with no problems, didn't even complain, but the new email was lost.

Checking the CPU, MB, drives, etc., temperatures show normal temperature accross the board:



The ambient temperature is 68 - 70 degree Fahrenheit...
The last two or three days, the system runs just fine, no issues whatsoever. That's not a guarantee, that the issue"s" won't return.

On the same system, new version of W10 Pro, 64-bit was installed, v2004/2H, on a different SSD. Nothing fancy, just pretty much default installation no additional software installed. The new installation also had a BSOD, but did not notice the exact cause, but had not been related to memory management. This installation is also running fine now.

Any idea what can cause the occasional BSODs and the system shutting itself off without warning?

Bad RAM connection maybe. Take out all the modules one by one and just put them back in if the contacts look okay. Maybe do the same for other cards/components too.

Start by going through this guide. MemTest86 - Test Results Memtest86 also checks the CPU. Most likely you have a bad CPU is why it is showing errors in the test.

All memory cards had been removed and they looked a ok. For some reason, from the beginning I suspected the CPU being bad. One of the reason is that MemTest86 did not detect/showed the "Lx" memory sizes:



CPUID does see it:



I tend to agree with "bro67", it's probably the CPU and not the memory.

Windows updates do install new drivers for the CPU. Is it possible these security updates cause these issues? After all, this system ran pretty much flawlessly for about eight years. Yes, aging hardware could cause this as well. Maybe the torture tests will have some more specific advise.

zbook said:

Please run:


GetMemoryConfiguration.bat - Click here to go to the BSOD batch repository to download and run this batch file.

Tuneup.bat - Click here to go to the BSOD batch repository to download and run this batch file.


V2 and DM: BSOD - Posting Instructions

Thanks z...

I did run both files as admin, the results are attached and all good... now....

Initially, I did not save the content of the tuneup in the command window. As such, the "tunaup.txt" file might be irrelevent. Some of the findings that I recall.

The "sfc /scannow" did detect integrity violation, that it could not fix.

The "dism /online /cleanup-image /scanhealth" detected component store corruption, that it deemed repearable online. The command "C:\WINDOWS\system32>dism /online /cleanup-image /restorehealth" was able to repair corruption.

The command "sfc /scannow" did not find integrity violation the second time around.

Thanks for the files, that did find and fixed issues. I did run the same commands manually about a month ego and there were no issues at that time.

In either case, before and and after repairing the store corruption, testing the memory with Prime95 fails with the same errors:

FATAL ERROR: Rounding was 0.498046875, expected less than 0.4
FATAL ERROR: Rounding was 0.5, expected less than 0.4

The memory test with Prime95 was with four threads, in all cases one or two threads failed pretty much instantly withouth completing the first run.

On the other hand testing the CPU only, that uses minimal memory, does not fail and completes the tests.

That seem to indicate that there is some bad block(s) on some of the memory cards. I cannot wait to replace the memory as the first step to remediate this issue...

I'll do it tomorrow... oh wait, it is tomorrow. I'll do it later today...

Night, night...

- - - Updated - - -

zbook said:

1) When available please upload V2 and DM:
BSOD - Posting Instructions
How to Upload and Post Screenshots and Files at Ten Forums

My AV didn't like the V2 log collector and removed it. I'll disable the AV and will run it again.

2) Upgrade the BIOS: F11 > F12
GA-Z77X-UP5 TH (rev. 1.0) Support | Motherboard - GIGABYTE Global

Updating the to F12 went fine, but cannot access the BIOS now:


Attachment 307646

The BIOS does not respond to the keyboard, regardless which keys or all keys are pressed; the keyboard worked just fine with BIOS v.F11. Once Windows 10 starts, the keybord has no issues.

The boot order changed from my "work" W10 to the test W10; the first boot device is a USB device by default. For the time being, the test SSD had been disconnected and will deal with it later.

3) For all tests / steps please post images into the thread.

For any problems posting images please use share links: one drive, drop box, or google drive

That won't work, I don't have any of those cloud accounts. I do have an iCloud account that's not used, I'll look if that can be used for sharing files.

4) Run memtest86+ version 5.31 for at least 8 passes.
Memtest86+ - Advanced Memory Diagnostic Tool
This may take hours so plan to run it overnight.

I forgot to run it overnight, but will let it run tonight. The initial, short test shows pretty much the same errors as the MemTest86 v.8.4:


Attachment 307647

5) Run: IPDT (typically takes < 15 min)
Download Intel(R) Processor Diagnostic Tool

The IPDT completed without errors with BIOS v. F11, relatively quickly, did not capture the results. With the F12 BIOS on the other hand, it is still running:



It's been than an hour or so, but no error, other than the CPU is running hot:



Normally, the CPU is close to 100 F degree max at ambient temps at 70F.

The system has been running just fine for the last couple of days...

zbook said:

The attachments displayed:

Code:

Message
Invalid Attachment specified. If you followed a valid link, please notify the administrator

There is a time factor when posting attachments.

Please post new attachments.

When available upload V2 results.

I've updated my previous post with the missing images...

The IPDT did complete with the F12 BIOS, without error:



I'll run the V2 script and upload the results, thanks for your help...