New System crashes while gaming

Toggle sidebar Toggle sidebar
Status
Not open for further replies.
D

Deleted member 2035586

Guest
Hi everyone,

I am usually a silent reader around here but recently I got problems with my new PC System myself. I moved from my old Phenom II 965 to the i5-6600k Skylake. Since then I get random crashes when playing (DotA 2, probably because I play it most of the time). The crashes seem random as they only happen every few days and I don't know if it is a hardware issue or by what component it is caused. My System is:


  • [new] i5-6600k (non-OCed)
    [new] Asus Z170-P
    [new] 2x8GB Corsair Vengence LPX 2400 (@2133 so also non-OCed)
    Powercolor PCS+ R9 290 (only factory OCed)
    XFX 650W Core Edition
    [new] 240 GB SanDisk Ultra II (Boot drive)
    256 GB Samsung 840 Evo
    [new] running Windows 10 Education edition
(new parts tagged)

What I tried so far:

  • Reset BIOS
    Change to old Graphics Card (HD6870) -> now randomly crashes at system startup instead of ingame
    Change to no graphics crad (no crashes in 4 day period)
    Checked RAM with memtest86+ -> 8 Passes in 14h no errors
    (Re)Installed Displaydrivers with DDU in safe mode
    Ran OCCT stability test overnight 12h CPU stresstest -> no crash
    Ran OCCT Power supply test 4h -> no crash (I dont think its the PSU since it handeled the old 125W Phenom just fine)
    Reinstalled Win 10
    Disabled fastboot
    Set Graphics to PCIe only in BIOS

I may have forgotten some things. Basically the errors seem completely unrelated since I got IRQL_NOT_LESS_OR_EQUAL, SYSTEM_THREAD_EXCEPTION_NOT_HANDELED and CLOCK_WATCHDOG_TIMEOUT in the past weeks, so wildy differing. The crashes never happened other than boot-time or when gaming, where the latter is much more prominent. The Crash ingame itself however is always identical: The image freezes and the sound (via HDM cable) is just some kind of noise (not like static however, more like a "brrrrrrrr"). In hopes of getting some information out of it I included the crashdump of the latest CLOCK_WATCHDOG_TIMEOUT: https://www.dropbox.com/s/kpiyxnyswxnlsp1/MEMORY.DMP?dl=0
(The upload could take a while -.-)

EDIT: Oh and temps are completely fine even under load I don't get over 70°C on the GPU and 55°C CPU. I also checked the windows event viewer and noticed I have literally thousands of errors over a couple of seconds (However not before/after every crash): "The attempt to connect to the RPCSS service was denied access for the COM Server application C:\Windows\Microsoft.NET\Framework64\v4.0.30319\NGenTask.exe to the user Unavailable\Unavailable SID (S-1-5-18) running in the application container Unavailable SID (Unavailable). The most likely cause is that the machine wide Access Limits do not grant the user or application local access permissions. The Access Limits can be modified using the Component Services administrative tool."
 
Sort by date Sort by votes
D

Deleted member 2035586

Guest


Just did that and there are four dump files (since win install erased all the older ones). Three of which occured with my old graphics card, cause by ntoskrnl.exe (x2) and FLTMGR.SYS. However, on the FLTMGR.SYS there are also NTFS.sys and ntoskrnl.exe highlighted in red on the bottom pane. On the most recent crash with my new graphics card it says it was caused by hal.dll, but ntoskrnl.exe is also again highlighted at the bottom in red. See screenshot here: http://imgur.com/SDArhJ9
 
Upvote 0 Downvote
B

BadAsAl

Distinguished
Oct 13, 2009
2,345
0
20,960
In that thread it looks like they tried Prime95 and it wouldn't even run. I see you already stress tested the CPU so I don't know if this will help.
The no crashing with only onboard graphics is interesting. Can you put in the R9 290 and set the BIOS to only use the addon graphics card and see if that helps?
 
Upvote 0 Downvote
D

Deleted member 2035586

Guest
Currently reinstalling all the drivers after DDU wipe. I already set the BIOS to only use PCIe Graphics the time I installed the Card, since I thought there could be a conflict with the iGPU. Sorry I forgot to mention that, basically I tried everything I could think of for the last month or so, so forgive me if I forget some smaller things. The crash you see where produces with Graphics set to PCIe specifically and I did not change the setting back to Auto.

EDIT: Oh and I never installed the gaming evolved app, that comes with the AMD DD, as I think it's bloatware... I don't need it, do I?

EDIT 2: From what I have read the ntoskrnl almost always points to faulty memory or storage (pagefile), so I am at this point considering another windows reinstall on my older Samsung SSD (which worked in several systems before) and another run of memtest86+ for a prolonged period of time (thinking of 24+ h).

EDIT 3: Just ran sfc/scannow and it found corrupted files, which the log identifies as the opencl.dll :
2016-03-08 15:23:20, Info CSI 000050ba [SR] Repairing corrupted file [l:23 ml:24]"\??\C:\Windows\SysWOW64"\[l:10]"opencl.dll" from store
 
Upvote 0 Downvote
D

Deleted member 2035586

Guest
Just got another freeze without any bluescreen and managed to loge the whole thing leading up to it in Hardwaremonitor:
https://docs.google.com/spreadsheets/d/1y3aSYbeFxNMyXOfckhteGGTi-rrtUrUNZBJQ4Dwfcok/edit?usp=sharing

I will look into using Seatools to check drive health. Sandisks own tool always displayed drive health as good. I really don't ant wo setup windows yet again. ugh -.-

EDIT: Seatools long term test turned up no error. Did another scf/scannow to see if maybe opencl got corrupted again but it also turned up no errors. Windows Eventviewer also displays nothing but the critical kernel power from me resetting the system after freezing. Since there was no BSOD Bluescreenview didn't turn up anything new either.

EDIT 2: After all this the crashes seem to happen more frequently and consitently. I still don't get bluescreens just freezes, which I have to do a reset from. But after I got several Freezes in a couple of hours there are there are still no events logged around the time of crash. I did find however, that every system start it seems to try and load the iGPU drivers (which are obviously not installed:
The igfxCUIService2.0.0.0 service failed to start due to the following error:
The system cannot find the file specified.
 
Upvote 0 Downvote
D

Deleted member 2035586

Guest
So for everyone still interested:
I updated the BIOS again two weeks before today, because ASUS had just released a new version. I hadn't had any Freezes for a long time and even thought it was fixed, but on friday last week the system froze again. I then proceeded to remove the new SSD from the System (unplugged SATA connector from motherboard) and reinstall Windows (yet again). It froze again and I am now going to send it in under warranty if possible. Since I am not sure which part is actually at fault I probably have to send in the whole lot...
 
Upvote 0 Downvote
D

Deleted member 2035586

Guest
So after sending everything back to the seller they replaced the RAM and it looks like it did the trick. The thing I take away from this is, that - apparently - 8 passes in 14h by Memtest86+ are by no means a guarantee fro good memory.
 
Upvote 0 Downvote
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom