Gigabyte Z97X-UD7 TH new build BSOD under GPU load - Windows 7 x64

[digikev]

Hi all, I am new here and new to building PC's. This is my first rig and having BSOD occurrences when the GPU is under load. The GPU is a Palit Nvidia GTX-970 with a single blower. It was bought in Amazon Warehouse Deals and is a previously used item. How much use is not fully known, but the deal wasn't significantly lower than full asking price. I have my concerns that it may be faulty. Before returning, I want to rule out any configuration problems or other software / hardware related issues first.

The MB BIOS has been flashed to F5b. This machine is also a Hackintosh with OS X Yosemite running on its own distinct HDD. Windows 7 Home Premium exists on another HDD.

I have updated all drivers in both OS's. OS X is also experiencing freezes after a period of time where the screen just locks, the HDD disk stops spinning and all functionality is disabled. It is in Win 7 when playing games the problem really comes to light. Monitor loses signal, HDD disk turns off and the system reboots. I have setup Win 7 to output dump files when BSOD occurs. Although I am unable to see the BSOD due to monitor signal loss, the dump file informs me that it occurred. The dump appears to point to the Nvidia GPU, but I lack expertise in assessing this information.

I have run 3DMark tests on the GPU, and this appeared to run OK with no BSOD taking place. However, GPU reach 74deg C and 3DMark returned an unspecified error (attached). I suspect the GPU is running much hotter under load of playing game as heat monitor registered 70deg C after reboot.

Shedding further light on this situation will be very welcome. Thank you in advance.

 

[digikev]

I followed the instructions for Driver Verifier after posting this thread and had to complete a system restore as when I booted after running Driver Verifier I had BSOD. I have attached the mini dump output from this, if it is any further help?

 

[Jared]

3: kd> .bugcheck
Bugcheck code 00000116
Arguments fffffa80`16ceb4e0 fffff880`04a352dc ffffffff`c000009a 00000000`00000004

3: kd> k
Child-SP          RetAddr           Call Site
fffff880`0611ea48 fffff880`04d01134 nt!KeBugCheckEx
fffff880`0611ea50 fffff880`04cd4867 dxgkrnl!TdrBugcheckOnTimeout+0xec
fffff880`0611ea90 fffff880`04d00f43 dxgkrnl!DXGADAPTER::Reset+0x2a3
fffff880`0611eb40 fffff880`04dd103d dxgkrnl!TdrResetFromTimeout+0x23
fffff880`0611ebc0 fffff800`02530b8a dxgmms1!VidSchiWorkerThread+0x101
fffff880`0611ec00 fffff800`022838e6 nt!PspSystemThreadStartup+0x5a
fffff880`0611ec40 00000000`00000000 nt!KxStartSystemThread+0x16

3: kd> u nvlddmkm+79c2dc
nvlddmkm+0x79c2dc:
fffff880`04a352dc 48ff257d97edff  jmp     qword ptr [nvlddmkm+0x675a60 (fffff880`0490ea60)]
fffff880`04a352e3 cc              int     3
fffff880`04a352e4 48ff255597edff  jmp     qword ptr [nvlddmkm+0x675a40 (fffff880`0490ea40)]
fffff880`04a352eb cc              int     3
fffff880`04a352ec 48ff255597edff  jmp     qword ptr [nvlddmkm+0x675a48 (fffff880`0490ea48)]
fffff880`04a352f3 cc              int     3
fffff880`04a352f4 48ff255597edff  jmp     qword ptr [nvlddmkm+0x675a50 (fffff880`0490ea50)]
fffff880`04a352fb cc              int     3

3: kd> lmvm nvlddmkm
start             end                 module name
fffff880`04299000 fffff880`04ca4000   nvlddmkm T (no symbols)           
    Loaded symbol image file: nvlddmkm.sys
    Image path: \SystemRoot\system32\DRIVERS\nvlddmkm.sys
    Image name: nvlddmkm.sys
    Timestamp:        Fri Mar 13 15:39:56 2015 (550304CC)
    CheckSum:         009CE879
    ImageSize:        00A0B000
    Translations:     0000.04b0 0000.04e4 0409.04b0 0409.04e4

So the Nvidia driver keeps crashign because it runs out of resources, the way I see it, there's two possibilities.

1. The driver is buggy, it's causing a memory leak of some kind.
2. The card is failing, the memory inside has developed a fault, thus preventing the card from functioning.

This driver version is consistent, so you could try rolling it back.

However, you say you've used a different OS, specifically OS X, and it is still crashing. Lastly, it only crashes on high loads.
All of this points towards a failing graphics card.

It was bought in Amazon Warehouse Deals and is a previously used item. How much use is not fully known, but the deal wasn't significantly lower than full asking price.

I would return it, get your money back and buy it brand new, that's your best option.

---

0: kd> .bugcheck
Bugcheck code 0000000A
Arguments fffff880`02e2f2b8 00000000`00000002 00000000`00000001 fffff800`022d58ac

0: kd> k
Child-SP          RetAddr           Call Site
fffff880`009a8308 fffff800`022cd469 nt!KeBugCheckEx
fffff880`009a8310 fffff800`022cc0e0 nt!KiBugCheckDispatch+0x69
fffff880`009a8450 fffff800`022d58ac nt!KiPageFault+0x260
fffff880`009a85e0 fffff800`027706e1 nt!KeWaitForSingleObject+0x13c
fffff880`009a8680 fffff880`02e298a9 nt!VerifierKeWaitForSingleObject+0x151
fffff880`009a8700 fffff880`02e2f2b8 prl_boot+0x48a9
fffff880`009a8708 fffff800`00000000 prl_boot+0xa2b8
fffff880`009a8710 00000000`00000001 0xfffff800`00000000
fffff880`009a8718 fffffa80`127b6b30 0x1
fffff880`009a8720 00000000`00000000 0xfffffa80`127b6b30

0: kd> .trap 0xfffff880009a8450
NOTE: The trap frame does not contain all registers.
Some register values may be zeroed or incorrect.
rax=000000000000024c rbx=0000000000000000 rcx=0000000000000002
rdx=0000000000000000 rsi=0000000000000000 rdi=0000000000000000
rip=fffff800022d58ac rsp=fffff880009a85e0 rbp=0000000000000000
 r8=fffff78000000008  r9=fffff88002e2f200 r10=0000000000000000
r11=fffff8000244ae80 r12=0000000000000000 r13=0000000000000000
r14=0000000000000000 r15=0000000000000000
iopl=0         nv up ei pl zr na po nc
nt!KeWaitForSingleObject+0x13c:
fffff800`022d58ac f00fba2e07      lock bts dword ptr [rsi],7 ds:00000000`00000000=????????

0: kd> lmvm prl_boot
start             end                 module name
fffff880`02e25000 fffff880`02e34000   prl_boot T (no symbols)           
    Loaded symbol image file: prl_boot.sys
    Image path: \SystemRoot\System32\Drivers\prl_boot.sys
    Image name: prl_boot.sys
    Timestamp:        Tue Feb 03 15:57:23 2015 (54D0EFE3)
    CheckSum:         00014927
    ImageSize:        0000F000
    Translations:     0000.04b0 0000.04e4 0409.04b0 0409.04e4

prl_boot.sys (Parallels Bootcamp Helper Driver (Part of Parallels Tools)) is causing issues.
It's waiting for an object which is part of pageable memory, not allowed.
Is this driver necessary? Do you need the software? If not, remove it.
If you do need it, then roll back, or update the driver as it's using poor programming techniques which is causing issues.

 

[digikev]

I appreciate your analysis, Jared. I suspected that it was a failing graphics card, but wanted confirmation before I send it back and find later it to be a config error on my part. I appreciate your time, thank you.

The Parallels Bootcamp Helper Driver allows me to run Win 7 from within Parallels Desktop app on OS X. However, I knew that it was damaged because it wasn't loading. After posting the BSOD dumps on this thread, I reconfigured Parallels and have it working again. I will keep a watchful eye on this and if any more errors occur attempt reinstalling Parallels or removing and forgoing the quick Windows access from OS X. Thanks again.