Windows 10: Random BSODs

Hi,

a few weeks ago i build a System from scratch with this specifications: 2x Xeon E5-2630 V4, Asus Z10PE-D8 WS Mainboard, 64 Gig ECC Ram, Geforce GTX 970, Soundblaster Z, Samsung SSD 950 Pro 512GB NVMe and a bequiet 650 Watt powersupply. After 2 weeks the System became unstable and the BSODs started accouring. Sometimes the System restarts fine for 2 days and runs smoothly. I ran CPU, GPU and Ram stresstests without any findings.
The gtx and the soundblaster are running fine in a different machine. No Errors there.
I reinstalled Windows without the graphicscard and the soundblaster and it seemed absolute stable but useless. So i installed Windows again with all the latest Drivers with my gc and soundcard and it ran for 2 days. Now the BSODs start to reappear when i restart the System. Usually a few seconds after Windows starts. I have no clue whats going on so any help is appreciated. I got the System gathering files attached. Cheers, Martin

:)

 

ntoskrnl.exe+1640e0 BSOD error

So far, with one memory chip no BSOD. But, getting a BSOD would be random and infrequent.

I've went through that article in the past and still have gotten BSOD.

 

DISM, Windows Reset/Refresh not working.

Faulty RAM can give rise to 'random' blue screens (BSODs).

 

All the BSOD dump files indicate a 0x124 WHEA (Windows Hardware Error Architecture) error. There is no further information in the dump files that I can find to help with the analysis. It is almost certainly hardware related but it could be a software utility program associated with the MB or graphics/sound cards attached.

Can you determine if it is the graphics card or the soundblaster card by running with only one at a time? That will help narrow it down.

 

What tests did you perform and what exactly were the results? We'd like to know details about the tests and like to see logs or screenshots/photos of the tests.

There are WER logs about bad dump files what is an indication towards faulty memory, this is another reason why details about the performed tests is important.
Code: Windows Error Reporting: +++ WER0 +++: Fehlerbucket BAD_DUMPFILE, Typ 0 Ereignisname: BlueScreen Antwort: Nicht verfügbar CAB-Datei-ID: 22eefd03-988d-4d0a-80bf-7a93ed017d57 Problemsignatur: P1: 124 P2: 0 P3: ffffe000d1a8c028 P4: fa08cdc0 P5: 20e0f P6: 10_0_10586 P7: 0_0 P8: 256_1 P9: P10: Angefügte Dateien: C:\Windows\Minidump\070516-6828-01.dmp C:\Windows\Temp\WER-8500-0.sysdata.xml C:\Windows\Temp\WER228B.tmp.WERInternalMetadata.xml Diese Dateien befinden sich möglicherweise hier: Analysesymbol: Es wird erneut nach einer Lösung gesucht: 1 Berichts-ID: 3c026cd6-2fa6-4deb-b16c-c9e342a4f539 Berichtstatus: 0 Bucket mit Hash:[/quote] Besides of that the dumps show a no-timeout error on a bus, I did not properly check which bus it is but likely the one closest to the memory.

 

@philc43 : i will give this a try again. Is there a program you know of that is starting/restarting the pc continously so i can see if it boots up stable?

@axe0 : i ran an Intel stress test program (forgot the Name) on the last Installation. Ram and CPU. I ran furrmark for the gpu. I did heavy fluidsims and 3d Rendering and the System never crashed on those occasions. Usally it just doesnt boot properly or gets a bsod a few seconds to a Minute after launching Windows.

Do you guys have any recomendations concerning Ram, Cpu and Gpu Tests?

Cheers, Martin

 

For furmark it is mainly the temperature you need to look at, not errors as I've only seen once an error from it.

Intel burntest I think? If so then it is the same as for furmark.

For the memory MemTest86+ is better to use.





Diagnostic Test

RAM TEST


Run MemTest86+ - Test RAM - Windows 10 Forums



Note



MemTest86+ needs to be run for at least 8 complete passes for conclusive results. Set it running before you go to bed and leave it overnight. We're looking for zero errors here. Even a single error will indicate RAM failure.

Make a photo of the result and post it.

Addition:
If errors show up you can stop the test, remove all sticks but 1 and test this single stick in each slot for 8 passes or until errors show, switch sticks and repeat.
If errors show up and you see them a lot later, no problem, the errors don't affect the test.

 

The temperatures were fine at all Tests. The gpu was never above 67° and the fan stayed at 39%. The same for both the cpus. They were way inside the tollerance. The 2 big noctua fans are doing their Job. I ll try to run the mem test next and run the System without soundcard and without grafixcard. thx for the help so far!