Unable to diagnose BSoD and various other crashes

[Gumbili]

Hiyas!

I'm hoping somebody is able to help me with a problem I'm currently trying to figure out. With my old graphics card (GTX 750 Ti) I occasionally have crashes while playing a somewhat graphics-intensive game (Final Fantasy XIV). The crashes are to desktop, of the program only. When I have my new gpu installed (XFX R9 680), I not only get these crashes more frequently, but also crashes where the screen goes black but the active programs keep running (can still talk to people in voice chat, sound still playing). On top of that, I'll sometimes have full system crashes (not BSoD). The longer the card is installed, the more frequent the crashes become, especially the more severe crashes. I've also recently (past two weeks or so) started getting BSoD crashes while using the GTX 750Ti, so far while playing Overwatch and Fortresscraft (a game that I assume is rather processor-intensive).

So far I've tried a fresh install of Windows 10, reinstalling drivers (with DDU, and used DDU for every driver update), running games with no additional programs active, killing all Raptr processes, and I bought a new PSU in the hopes that the issue was power-related (my old PSU was an unused hand-me-down, but no clue how old). Obviously none of these has worked. I'm not very competent at diagnosing the problem, and google has only gotten me so far, so I'd be very grateful for any input! What can I do to identify the problem area? What additional information can I provide to assist in figuring out the issue?

Thank you so much in advance for any help anyone can provide!

PC info:

Processor: Intel i5-4670K 3.40GHz
GPU: GTX 750 Ti OR XFX R9 680
PSU: EVGA SuperNOVA 750 P2
Mobo: Asus Hero VI Maximus
OS: Windows 10 Pro 64 bit


Blue Screen Viewer info:

Spoiler

==================================================
Dump File : 101916-3890-01.dmp
Crash Time : 10/19/2016 6:34:27 PM
Bug Check String : PFN_LIST_CORRUPT
Bug Check Code : 0x0000004e
Parameter 1 : 00000000`00000002
Parameter 2 : 00000000`001632cd
Parameter 3 : 00000000`0041efff
Parameter 4 : 00000000`00000001
Caused By Driver : ntoskrnl.exe
Caused By Address : ntoskrnl.exe+14a2c0
File Description :
Product Name :
Company :
File Version :
Processor : x64
Crash Address : ntoskrnl.exe+14a2c0
Stack Address 1 :
Stack Address 2 :
Stack Address 3 :
Computer Name :
Full Path : C:\Windows\Minidump\101916-3890-01.dmp
Processors Count : 4
Major Version : 15
Minor Version : 14393
Dump File Size : 528,468
Dump File Time : 10/19/2016 6:35:02 PM
==================================================

==================================================
Dump File : 101016-3859-01.dmp
Crash Time : 10/10/2016 7:43:00 PM
Bug Check String : PFN_LIST_CORRUPT
Bug Check Code : 0x0000004e
Parameter 1 : 00000000`00000002
Parameter 2 : 00000000`0033d768
Parameter 3 : 00000000`0041efff
Parameter 4 : 00000000`00000001
Caused By Driver : FLTMGR.SYS
Caused By Address : FLTMGR.SYS+485b
File Description :
Product Name :
Company :
File Version :
Processor : x64
Crash Address : ntoskrnl.exe+14a2b0
Stack Address 1 :
Stack Address 2 :
Stack Address 3 :
Computer Name :
Full Path : C:\Windows\Minidump\101016-3859-01.dmp
Processors Count : 4
Major Version : 15
Minor Version : 14393
Dump File Size : 533,644
Dump File Time : 10/10/2016 7:43:34 PM
==================================================

==================================================
Dump File : 101016-3765-01.dmp
Crash Time : 10/10/2016 1:53:02 AM
Bug Check String : IRQL_NOT_LESS_OR_EQUAL
Bug Check Code : 0x0000000a
Parameter 1 : ffff8980`006e4900
Parameter 2 : 00000000`00000002
Parameter 3 : 00000000`00000000
Parameter 4 : fffff801`7782c196
Caused By Driver : ntoskrnl.exe
Caused By Address : ntoskrnl.exe+14a2b0
File Description :
Product Name :
Company :
File Version :
Processor : x64
Crash Address : ntoskrnl.exe+14a2b0
Stack Address 1 :
Stack Address 2 :
Stack Address 3 :
Computer Name :
Full Path : C:\Windows\Minidump\101016-3765-01.dmp
Processors Count : 4
Major Version : 15
Minor Version : 14393
Dump File Size : 481,196
Dump File Time : 10/10/2016 1:53:33 AM
==================================================

==================================================
Dump File : 100916-3796-01.dmp
Crash Time : 10/9/2016 10:03:15 PM
Bug Check String : IRQL_NOT_LESS_OR_EQUAL
Bug Check Code : 0x0000000a
Parameter 1 : fffffb00`00923900
Parameter 2 : 00000000`00000002
Parameter 3 : 00000000`00000000
Parameter 4 : fffff803`7461b196
Caused By Driver : ntoskrnl.exe
Caused By Address : ntoskrnl.exe+14a2b0
File Description :
Product Name :
Company :
File Version :
Processor : x64
Crash Address : ntoskrnl.exe+14a2b0
Stack Address 1 :
Stack Address 2 :
Stack Address 3 :
Computer Name :
Full Path : C:\Windows\Minidump\100916-3796-01.dmp
Processors Count : 4
Major Version : 15
Minor Version : 14393
Dump File Size : 477,660
Dump File Time : 10/9/2016 10:03:46 PM
==================================================

I also pulled the .dmp files from these, but I don't know if I can attach them here or how to provide that info.

 

[Ashbacher7]

Try updating BIOS, and in the start menu, typing in chkdsk.exe and let that run and see if that fixes your issue.

 

[panathas]

It's either the drive or RAM. Do you have a hard drive or an SSD? Also what RAM do you have? You haven't mentioned them in the specs. Finally are you sure that your CPU and GPU aren't overheating?

 

[Gumbili]

It's either the drive or RAM. Do you have a hard drive or an SSD? Also what RAM do you have? You haven't mentioned them in the specs. Finally are you sure that your CPU and GPU aren't overheating?

Sorry I left those out of the specs!
RAM: 2x Crucial Ballistix Sport 8GB DDR3 1600 CL9 1.5V (running in single because my PC doesn't boot if I have a stick in the mobo's 1 and/or 2 slots)
Drives: Samsung SSD 850 EVO 250GB
WDC WD5002AALX-00J37A0 500GB (not SSD)

I also forgot to mention that I've tried running the PC with each RAM stick on its own in both 3 and 4 ports (as the 1 and 2 slots don't work) with the crashes still occurring regardless of RAM stick and location.

Neither CPU nor GPU gets very hot at all - I have solid airflow in my case. I don't remember the exact temperature numbers from the last time I monitored under heavy load, but it wasn't a large uptick from passive temperature levels and certainly not overheating.

 

[Gumbili]

Try updating BIOS, and in the start menu, typing in chkdsk.exe and let that run and see if that fixes your issue.

I flashed the most recent BIOS maybe a month ago. When I run chkdsk.exe a window opens for a fraction of a second then disappears - I'm not sure if this is running in the background or if something's not working, but after waiting for ~20min, I haven't seen any indication that it's run or completed. I'm just sort of guessing that it's the conhost.exe that runs this, and the same thing occurs if I run conhost, either through Run or its application icon, including in administrator. Googling now for a solution.

Edit: ran chkdsk /f on both drives, I'll see if any crashes still occur. Thank you both for your help!

 

[panathas]

The fact that you can't run the RAM in dual channel mode is strange and makes me think that you should test your RAM further with memtest86+. That's not normal and something must be going with your RAM or your motherboard.

If you run chkdsk and it doesn't report any issues then your drives are most likely fine. Since you have a Samsung SSD you should be using Samsung magician. It reports back the SSD health. You could also use the CrystalDiskInfo program. It reads the smart status of your drivers (SSD & HDD) and reports back their health status. If you don't have it, you should download it.

Finally after reading the dumb files, you posted, it seems you are also having driver issues. Make sure you have the latest drivers for everything not just the GPU.

 

[Gumbili]

The fact that you can't run the RAM in dual channel mode is strange and makes me think that you should test your RAM further with memtest86+. That's not normal and something must be going with your RAM or your motherboard.

If you run chkdsk and it doesn't report any issues then your drives are most likely fine. Since you have a Samsung SSD you should be using Samsung magician. It reports back the SSD health. You could also use the CrystalDiskInfo program. It reads the smart status of your drivers (SSD & HDD) and reports back their health status. If you don't have it, you should download it.

Finally after reading the dumb files, you posted, it seems you are also having driver issues. Make sure you have the latest drivers for everything not just the GPU.

I don't know if it's dual channel that's actually the issue - it's the 1 and 2 slots specifically that cause my computer to not boot, even with a single stick.

chkdsk, Magician, and CrystalDiskInfo all report good health status for both drives (thank you for the tip on getting CrystalDiskInfo!).

Regarding the driver issues, I have the most up-to-date GPU and motherboard drivers, and device manager doesn't report any drivers not working properly. Are there any I would be missing? I don't have any extra hardware like a wireless card, and Magician reports no new firmware updates for my SSD. I had googled the BSoD codes from the earlier crashes and saw that it was driver-related, but I assumed that was related to the Radeon card. Now that I've gotten them with the Nvidia card as well, I suppose that means it's not the case...

I started running memtest86+ but had to interrupt it after the first pass due to work. The first pass reported 0 errors, and I intend to run the full test tonight (would there be any issue with leaving it to run overnight?).

Thank you so much again for your continued help, I appreciate it immensely!

Update: ran memtest86+ (4 passes) with 0 errors. Could you suggest any next steps? If I understand what you've written, it sounds like the issue is either with the motherboard, some drivers, or both.

 

[panathas]

OK. Your drives and RAM are fine.

The fact that your RAM doesn't work on slots 1 & 2 is strange and maybe you have damaged hardware somewhere. Either the motherboard slots 1 & 2 got somehow dead or your CPU socket has some bend pins, and those specific pins are the ones that connect the CPU to those 2 slots. You could try re-seating the CPU and inspecting the CPU socket for bend pins but you should know what you are doing. I don't know if this issue is related or not with the crashes but it exists and I think that you should fix it.

If your motherboard is still under warranty and is indeed damaged you should RMA it. Maybe you should take your system at a repair shop or have a technician take a look at it.

Did you assemble this system yourself? Also what cooler do you have? I am asking because some CPU coolers cause a lot of stress on the CPU socket, during the installation process and if you aren't very careful you could damage the board and lose some motherboard memory slots.

As for the driver issue. There are motherboard drivers but there are also sound, LAN, USB3, extra sata3 chip etc. When you said motherboard drivers, did you update those too? The thing about those drivers is that motherboard makers rarely update the motherboard pages at their sites and when they do they only update their own software leaving out other motherboard devices that are produced by other manufacturers. This problem gets even greater when a new OS appears, just like it's happening right now with win10. Usually motherboard makers just put the initial versions on their sites and most of those have bugs. Also the drivers that win10 automatically installs are generic ones that may or may not have bugs. So you have to search for each specific driver manually. You have to search your board's manual and find the manufacturer and the specific model of your LAN, audio chip, extra USB3 or sata chipsets and then go to those sites and search for the specific win10 drivers. Also you have to go at Intel's site and download the latest win10 Intel rapid storage driver and the win10 chipset driver for your motherboard's specific chipset. You can find all of those model names at your board's manual.

Sometimes the driver thing is a mess and one bad driver can make your systems life like hell. But keep in mind that you may not have a driver issue but a hardware issue if you motherboard is indeed damaged.

 

[veldrane2]

Also, just FYI, do not use DDU on future re-installs of drivers. DDU is outdated and does not keep up with Windows 10 versions and updates. It removes stuff it shouldn't touch.

If you want to clean uninstall drivers, then go control panel. programs features, uninstall from there, then download the AMD clean-up utility from AMD, and use that, then manually delete any applicable directories, then run CCLeaner to get rid of left overs in the registry. Thats for the AMD drivers. I do not know the equivalent of a clean-up utility for Nvidia drivers.

Since you have 2 GPUs listed in your system specs, and possibly have had drivers for both ? You may have some leftover conflicts.

That's just drivers. But as stated above, if your mobo doesn't read RAM in slots 1 and 2, then I simply wouldn't trust it hardware wise.

 

[Gumbili]

OK. Your drives and RAM are fine.

The fact that your RAM doesn't work on slots 1 & 2 is strange and maybe you have damaged hardware somewhere. Either the motherboard slots 1 & 2 got somehow dead or your CPU socket has some bend pins, and those specific pins are the ones that connect the CPU to those 2 slots. You could try re-seating the CPU and inspecting the CPU socket for bend pins but you should know what you are doing. I don't know if this issue is related or not with the crashes but it exists and I think that you should fix it.

If your motherboard is still under warranty and is indeed damaged you should RMA it. Maybe you should take your system at a repair shop or have a technician take a look at it.

Did you assemble this system yourself? Also what cooler do you have? I am asking because some CPU coolers cause a lot of stress on the CPU socket, during the installation process and if you aren't very careful you could damage the board and lose some motherboard memory slots.

As for the driver issue. There are motherboard drivers but there are also sound, LAN, USB3, extra sata3 chip etc. When you said motherboard drivers, did you update those too? The thing about those drivers is that motherboard makers rarely update the motherboard pages at their sites and when they do they only update their own software leaving out other motherboard devices that are produced by other manufacturers. This problem gets even greater when a new OS appears, just like it's happening right now with win10. Usually motherboard makers just put the initial versions on their sites and most of those have bugs. Also the drivers that win10 automatically installs are generic ones that may or may not have bugs. So you have to search for each specific driver manually. You have to search your board's manual and find the manufacturer and the specific model of your LAN, audio chip, extra USB3 or sata chipsets and then go to those sites and search for the specific win10 drivers. Also you have to go at Intel's site and download the latest win10 Intel rapid storage driver and the win10 chipset driver for your motherboard's specific chipset. You can find all of those model names at your board's manual.

Sometimes the driver thing is a mess and one bad driver can make your systems life like hell. But keep in mind that you may not have a driver issue but a hardware issue if you motherboard is indeed damaged.

Thanks so much for all the info, going to hunt down all the drivers ASAP and reseat my CPU / check the socket. I was using the manufacturer's website for them, aside from the chipset drivers (thanks to a random google tangent suggesting that).

I did assemble my system myself, cooler is a Deepcool Gammax 400, though the slots didn't function prior to installing that. I was using some Intel heatsink before, can't remember what specifically it was for the life of me. I did check the socket when reseating my CPU with the new cooler, but it's certainly possible I missed something. I'm no longer under warranty for my motherboard, and would have RMA'd it but I didn't discover the problem with the 1 and 2 slots until relatively recently when I upgraded from a single stick of 8G RAM to my current memory.

Also, thank you veldrane2 for the tip about DDU, I'll hunt down a similar method for Nvidia drivers and get rid of everything before reinstalling.