off-topic: frequent PC crashes / BSOD - any help?

I hope you don't mind asking this question in this forum. But since 2 days, my PC frequently crashes with a BSOD (and I did not install any new software/driver or hardware).

The Stop Code WHEA_UNCORRECTABLE_ERROR (0x124) is related to a hardware failure, but I did not find out which component might be broken. I can only say for sure it's not the memory because the memory test program said they're fine.

My guess is that it might be the video card because it seems that the PC only crashes in 2D mode, not in 3D. With that said, I can play the games, but e.g. as soon as I exit AC (the game engine) to the main menu, the PC crashes. Same e.g. with Motorsport Manager. No problem while gaming, but as soon as I exit (or shortly after), BSOD.

I wanted to try out an older video card to isolate the problem, but I only have an old GTX 570 here. So when I installed this one, I had different problems (occasional "black screen"/signal lost). Maybe it's not fully compatible with modern mobos or it simply wasn't capable to handle WQHD.

So before I go to the store next week, I was hoping somebody else has some ides. Is there a way to test the video card (GTX 1080) for failures?

I also updated drivers and all thats stuff now, but the problems remain the same.

Last but not least, I noticed that if your PC crashes right after exiting AC and you made a pb, your personalbest.ini will be empty afterwards

Edit: The memory dump didn't really help me. All I know it doesn't have to be the CPU only because it's telling GenuineIntel.

BugCheck 124, {0, ffff8b83a7518028, bf800000, 114}
Page e9474 not present in the dump file. Type ".hh dbgerr004" for details
Page 400 not present in the dump file. Type ".hh dbgerr004" for details
Page 400 not present in the dump file. Type ".hh dbgerr004" for details
Page 400 not present in the dump file. Type ".hh dbgerr004" for details
Probably caused by : GenuineIntel

Edit2: Just in case there's an IT Pro here... some additional debugging information:

4: kd> !errrec ffff8b83a7518028
===============================================================================
Common Platform Error Record @ ffff8b83a7518028
-------------------------------------------------------------------------------
Record Id : 01d2b5c6d61944e4
Severity : Fatal (1)
Length : 928
Creator : Microsoft
Notify Type : Machine Check Exception
Timestamp : 4/15/2017 9:18:53 (UTC)
Flags : 0x00000000
===============================================================================
Section 0 : Processor Generic
-------------------------------------------------------------------------------
Descriptor @ ffff8b83a75180a8
Section @ ffff8b83a7518180
Offset : 344
Length : 192
Flags : 0x00000001 Primary
Severity : Fatal
Proc. Type : x86/x64
Instr. Set : x64
Error Type : Cache error
Operation : Generic
Flags : 0x00
Level : 0
CPU Version : 0x00000000000506e3
Processor ID : 0x0000000000000004
===============================================================================
Section 1 : x86/x64 Processor Specific
-------------------------------------------------------------------------------
Descriptor @ ffff8b83a75180f0
Section @ ffff8b83a7518240
Offset : 536
Length : 128
Flags : 0x00000000
Severity : Fatal
Local APIC Id : 0x0000000000000004
CPU Id : e3 06 05 00 00 08 10 04 - bf fb fa 7f ff fb eb bf
00 00 00 00 00 00 00 00 - 00 00 00 00 00 00 00 00
00 00 00 00 00 00 00 00 - 00 00 00 00 00 00 00 00
Proc. Info 0 @ ffff8b83a7518240
===============================================================================
Section 2 : x86/x64 MCA
-------------------------------------------------------------------------------
Descriptor @ ffff8b83a7518138
Section @ ffff8b83a75182c0
Offset : 664
Length : 264
Flags : 0x00000000
Severity : Fatal
Error : DCACHEL0_RD_ERR (Proc 4 Bank 1)
Status : 0xbf80000000000114
Address : 0x000000001d5aa780
Misc. : 0x0000000000000086

 

I wish I could help you, but unfortunately I have no clue. , good luck I hope it can be fixed soon.

 

Actually running the Intel Processor Diagnostic Tool (passed). The good thing is that I still have warranty on the components, but I need to be able to prove that one of these components really failed.

 

First just because you havent updated any drivers, doesnt mean that windows has not done so sneakily.

Fully remove all GPU drivers and replacd with newest from nVidia. Also check you have all latest motherboard drivers etc.

If you want to really test your CPU try this:

http://www.majorgeeks.com/files/details/intelburntest.html

Works for AMD cpu too btw! Finds issues with overclocked cpu when hours of prime95 thinks its fine... so imho a very good test.

Also when you say your RAM passed the test, did you boot off a usb stick into memtest 86 and run for several hours? Otherwise whatever test you used may not pick up the faulty RAM stick. Any memtest run in windows is not reliable test.

Good luck finding the issue Ray!

Edit! Most inportantly when troubleshooting start at the bottom of the OSI model - PHYSICAL LAYER!

So my first advice would be to open PC, remove of dust, disconnect and reconnect every connection including removing and re-seating the RAM and video card (not CPU tho lol!)

Many times I have solved repeated BSOD by re-seating the video card. Especially if your PC doesnt lie down normally, the GPU sags out the slot slightly causing BSOD. In my PC i had to eventually move it to another PCI Express slot in the end as it was sagging too much repeatedly causing BSOD.

Once you re seated it try leaving PC lying down and see if it still crashes as this way the GPU simply cannot sag out the slot.

 

Thank you for your feedback, @LeiF

Well, I'm aware of automatic updates and stuff like this but(I don't let it do automatically though (as far as possible) - just as a side note, I work in IT, but I'm more system admin than engineer.

What I did so far in the last 2 days:
- I opened the case, removed the video card and installed it again in the same slot
- I switched the two memory DIMM. I could still try only using one DIMM...maybe worth a try.
- I removed "unnecessary" USB devices
- checked memory with mdsched.exe (part of Windows 10, it's booting and doing the test w/o loading the OS). I only run the default test yet (approx. 30 minutes), not the advanced. So I may do the intense one. I did not use memtest86 this time (simply don't have a working USB stick at home...both seem broken). I may do a longer test when I won't use the PC. At least gaming is working *g* It simply crashes afterwards.

- I updated Chipset and video drivers, but I did not yet update the firmware (as mentioned, no USB stick.
- I even intended to install the creators updates, but when I even had BSOD during the download, I thought it would be a very bad idea to try an OS update

Re-seating the GPU doesn't seem possible. I thought only the first slot has the full power? At least some IT guy once told me to always use the first slot for the GPU. Or is this old info before SLI become popular again? The Asus Maximus Ranger has 2 slots, so I could try the other one - but I wouldn't be able to use the soundcard anymore in that case (space problem).

Thanks for the link for the CPU test. I'll test it. I did not overclock my CPU. The debugging tool told me it's running at 4008 MHz though while the maximum is 4000. But I don't know if this can be called overclocked.

Anyway, I think I'll have a break now. Don't want to deal around all the time with that PC stuff (this is already annoying enough sometimes in the office We all know Murphy).

 

Now this is interesting....
Just ran the CPU stress test you mentioned. That worked fine and I decided to stop it after 5 cycles (after 45 minutes). A second after I hit stop....BSOD. Looks like my PC doesn't like to be NOT stressed. Like I said, also gaming is no problem, but as soon as I exit the game - BSOD.

 

After break you can try some (very) old GPU drivers. Also try turn off all power saving and automatic update things (all, no only GPU related).
Edit: BTW I have some memory problem with my old 760 and Nvidia driver sometime crash while video playing, but while gaming no problems. It´s hardware issue, but some drivers make it less noticeable.

 

Regarding GPU in different slot - as long as it is the same number of PCI-x lanes I have found that it is same performance in my tests. Motherboard dependent of course.

Now that is weird, so it crashes after cpu stops being used! That test is not graphical in any way so points away from a GPU related issue. Oh well at least it isnt crashing during races!

Good luck once again!

 

Most probably I'll go to the store on Tuesday and let the engineer have a look at it. It's pretty strange and I'm running out of ideas. If I play a game and exit it normally...BSOD in most caes. If I play a game, due alt-tab and end it with Task Manager, no BSOD (so far). But it might just be a hazard.
I suppose I'll anyway run the long memory test later on.

Like you said, at least I can run AC, but it's no fun if the system isn't reliable anymore. Also, exiting AC with task manager means that no pb will be saved anymore.

 

Your Power Supply (PSU) may be going bad. I've had it happen myself. Some of the weirdest things can trigger a PSU BSOD. If you've got access to an equal or higher watt PSU you might give that a try to check.

You might also check to see if it can ventilate from the bottom well if that's where its mounted. Mine attracts dust like a vacuum cleaner and after a while the fan runs constantly and it gets scalding hot.

 

Thank you for this hint. I never thought of the PSU, but it's indeed the only component I did not change (aside the case) when I built the new system. So who knows....
Mine is not dusty and the fan is running fine (and it's still cool) but I'll keep it in mind and tell 'em when I go to the technical support.

 

Maybe it anyway was somehow software related. After the BIOS update which of course reset all settings to default, I anyway had to re-build the RAID. So I did a fresh install of Windows 10 to the latest build (data were anyway on another partition/disk). Just tried AC (and some other stuff that caused a BSOD before) and I don't get a BSOD anymore after exiting the game.

 

Even though the memory test is fine if you can I'd run with the memory modules individually. Most of the new build PC issues I've had have been caused by faulty/incompatible ram. Sometimes even swapping them over makes a difference...

 

I already swapped them two days ago ;-) But yes, I can do that test as long as I have warranty. But I'm pretty confident that the fresh install solved my problem.

 

Just in case somebody has a similiar problem once. After analyzing the memory dump and some research, I found out that it's a Level 0 Cache read error (DCACHEL0_RD_ERR (Proc 4 Bank 1). The vendor replaced the CPU under warranty after they also run some tests.

So if you ever have to debug a memory dump:
https://bsodtutorials.wordpress.com/2015/08/07/the-complete-debugging-guide-to-stop-0x124-part-3/