Gigabyte Aorus GA-Z270X-Gaming K5 - Crashing under GFX Load

Komedy · 7 Mar 2017 at 07:40

I had previously believed the issue i'll describe now was limited to World of Warcraft however now turns out to be more widespread.

I have built the following system as a little stop gap using a previous GTX680, everything else is new.

Gigabyte GA-Z270X - Gaming 5
Corsair 8GB DDR4 @ 2400MHz
G4560 3.50GHz
GTX680
EVGA 650w Gold PSU

After a fresh install of Windows 10 and on attempting to play WoW the PC would after a few mins of loading the game crash. Its a hard crashes that requires a reboot. No recovery.

In order to isolate the problem I have undertaken the following steps and so far none have resolved the issue.

Replaced the GTX680 with an HD7970 (full clean install of Windows at same time)
Replaced the PSU with a 900w unit
Updated the BIOS to F4
Run Memtest with no errors reported
Ensured I'm running the latest Intel Chipset Drivers
Ensured I'm using the latest Nvidia Drivers
Tried installation of Windows to a different Drive

So right now after yet another full install of Windows I have a pretty stripped back installation, all other hard drives out with only one SSD running Windows. No other programs installed other than the following bench marks.

Prime95 - Runs with no problems and during runs the CPU does not exceed 45oC
Heaven Benchmark - Crashes after a few seconds. GPU temp at time of crash 50oC
3dMark - Cloudflare - Completes the demo and benchmarks
3dMark - Skydiver - Completes the demo but crashes a few seconds into the 1st benchmark
3dMark - Timespy - Crashes

So now I am suspecting that the motherboard is faulty. Before I begin the RMA process can anyone suggest anything to try.

CapitalOne · 7 Mar 2017 at 07:43

Anything in the windows eveny log?

Komedy · 7 Mar 2017 at 07:51

CapitalOne said:
Anything in the windows eveny log?

Not looked in there. After each crash its a total freeze, black screen etc requiring a full reboot into Windows

CapitalOne · 7 Mar 2017 at 07:57

Take a look in the system log when it boots back up. Maybe some clues.

Komedy · 7 Mar 2017 at 08:07

CapitalOne said:
Take a look in the system log when it boots back up. Maybe some clues.

Okies will have a look in there this evening

orbitalwalsh · 7 Mar 2017 at 08:40

Tried running without GPU and using onboard graphics ? One isolation test i saw you didn't perform and run all the same tests

Komedy · 7 Mar 2017 at 09:50

orbitalwalsh said:
Tried running without GPU and using onboard graphics ? One isolation test i saw you didn't perform and run all the same tests

Yes forgot to add that - I removed the GPU and ran with the Intel IGPU and while WoW didnt crash it was unplayable at under 5fps.

Did not run any of the other benches though.

Will also add that WoW is running completely untouched with no addons at all.

hornetstinger · 7 Mar 2017 at 09:55

Try another GPU you know is working. If it crashes then pointing to PSU.
Try another PSU with your GTX680, if it doesn't crash then it's your EVGA 650w Gold PSU.#

Try underspeccing the memory, if it's CL9 RAM, set it to CL10 2200MHz

andicole0 · 7 Mar 2017 at 09:58

What are your temps like both CPU and GPU, it might need a heatsink refitting.
Andi.

Komedy · 7 Mar 2017 at 10:04

hornetstinger said:
Try another GPU you know is working. If it crashes then pointing to PSU.
Try another PSU with your GTX680, if it doesn't crash then it's your EVGA 650w Gold PSU.#

Try underspeccing the memory, if it's CL9 RAM, set it to CL10 2200MHz

I know the GPUs are both working as have tried them in my sons machine. I get the crash no matter which GPU is in

I have tried another PSU - a 900w PSU that I know is working

Komedy · 7 Mar 2017 at 10:04

andicole0 said:
What are your temps like both CPU and GPU, it might need a heatsink refitting.
Andi.

In Prime95 the CPU tops at 45oC

In Heaven Bench Mark the GPU reports 50oC just before the hard crash

orbitalwalsh · 7 Mar 2017 at 10:21

Komedy said:
Yes forgot to add that - I removed the GPU and ran with the Intel IGPU and while WoW didnt crash it was unplayable at under 5fps.

Did not run any of the other benches though.

Will also add that WoW is running completely untouched with no addons at all.

worth trying other benches, yes you'll get super low frames but just checking it makes it through without the hard crashing your getting . Least then you've ruled out CPU, its socket and its power delivery . As stated with ram, would be running it at Intel default of 2133 for the purpose of trouble shooting.

also, tried GPU in second PCIe 16x slot?

again just trying to rule out any connection issues

if you do have to RMA, post up progress and end report

Komedy · 7 Mar 2017 at 10:34

I'll give all these suggestions a run this evening and report back

orbitalwalsh · 7 Mar 2017 at 12:55

Komedy said:
I'll give all these suggestions a run this evening and report back

last ditch attempt after all that.

"might sound silly but they might want to check the CPU pins on the board too. Sometimes if they are not making full contact (one is slightly bent for example) it can cuase issues just like that." was advised to send you that

Komedy · 7 Mar 2017 at 20:29

Ok so back home and doing some of the stuff suggested above will add as I go along.

Tried another Heaven Bench Mark and this time ran for a few minutes before crashing. Temp reported as 50oC however crashed with a hard lock just before the end.
Event Viewer shows under critical errors : Event ID 41 Kernel-Power
Ran XTU Stress Test on CPU - Max Temp 42oC
Mem Set to BIOS Default 2133
So card in next slot was a disaster. Cant even get loading into windows without it freezing and crashing out. On a second attempt again uninstalling and installing drivers it caused issues at the bios loading screen and again froze on loading the desktop.
Removed the 680 and have run all benchmarks with the IGPU and while scores are pretty kak all complete
Popped the 680 into my sons i5 7600k and its running benches without a hiccup

orbitalwalsh · 8 Mar 2017 at 10:44

Komedy said:
Ok so back home and doing some of the stuff suggested above will add as I go along.

Tried another Heaven Bench Mark and this time ran for a few minutes before crashing. Temp reported as 50oC however crashed with a hard lock just before the end.

Event Viewer shows under critical errors : Event ID 41 Kernel-Power

Ran XTU Stress Test on CPU - Max Temp 42oC

Mem Set to BIOS Default 2133

So card in next slot was a disaster. Cant even get loading into windows without it freezing and crashing out. On a second attempt again uninstalling and installing drivers it caused issues at the bios loading screen and again froze on loading the desktop.

Removed the 680 and have run all benchmarks with the IGPU and while scores are pretty kak all complete

Popped the 680 into my sons i5 7600k and its running benches without a hiccup

damn, sorry to hear that. Have you tried reseating CPU? bit of a pain but uneven pin contact can cause issues, seems 85% of RMAs is this issue

and im guessing it the PCIe should be default Gen. 3.0 but worth a check to see if it's somehow on Gen 2.0- only reason why i can think of Windows not paying ball on 2nd slot.

Board from OCUK ? should have a speedy RMA turn around

most likely recommend get that going and try a CPU resit as well as you have to that the board out so might as well spend an extra few mins doing it. if it works you can cancel, if it doesn't, board would be sent off next day

Komedy · 9 Mar 2017 at 07:11

its out and winging its way back to OCuK

orbitalwalsh · 9 Mar 2017 at 08:25

Komedy said:
its out and winging its way back to OCuK

shame to here, but OCUK and Gigabyte UK normally have a quick RMA turn around

keep us updated to how it goes

Darren_uk · 13 Mar 2017 at 14:23

http://www.tomshardware.co.uk/faq/id-3128570/fix-windows-error-event-kernel-power.html

https://www.google.co.uk/webhp?sourceid=chrome-instant&ion=1&espv=2&ie=UTF-8#q=Event+ID+41&*