Cant find out why PC shutdowns during gaming

[HunkUmbrella]

Hey guys, I am tired and desperate to find out what the problem makes this happen.
No errors, blue screens, driver crashes, it just suddenly shut downs and restart.
I got:
motherboard - Z97X SocForce, bios ver F7
CPU - Intel I7 4790K cooling by Zalman CNPS 10X Extreme
GPU -GV-N980G1Gaming-4GD/gtx 980 rev 1.0 powered by 2 x 8 pin - 8 pin
RAM -16Gb DDR-III 1600MHz Kingston HyperX (KHX16C9K2/16) 2*8KIT   
PSU - Ice Hammer IH - 1600W
Win 7 Ultimate x64

Most interesting part that this is happening not always but often, not in all games, but I noticed it happens much often when I play games on high settings, but strangly it somehow occur not on all, for example:
- BF4 on ultra - stable - but Tropico 5 on high - shutsdown
ARMA 3 - very high - stable most of time - War Thunder - high settings - shutdowns often
CoH 2 - ultra, most stable, but can shutdown.

Now what I've tried, I visited my service center, we tested my GTX 980 on overheating, then my RAM, I tested my RAM 2 times with Memtest at home, I showed up my motherboard, showed my PSU, all were fine.
I upgraded motherboard bios using official @bios and bios versions on Gigabyte official support site, but that shutdown during gaming problem remained!
I though it was GPU overheating, and you know cooling helped a bit, it becomes not so often, but not in GPU were the problem, cause we tested it on another machine and all was fine, during hours and hours of hardcore graphic gaming.
Then I started to suspect my PSU again, cause I lowered GPU stats to make it more cooler and stable with 100% fans working, but then I realized, that the problem is not in PSU, cause I could play the games that shut downs on ultras for hours if the air was cooled.
CPU overheating then? seems not, but I can believe in everything now! I even started to think that when GPU starting to work hard, it overheat some motherboard details that are near, for example bridge or something else.
Then I thought that maybe the problem is not in hardware, but software, so I updated bios again to final version, checked all my drivers and etc, and still I failed, reinstalled Win7,re install all drivers cleanly but problems still remains.
So I need your help, what it can be!?
Interesting things that i noticed, that:
Bios runs I7 4790k to 4,40ghz (without Turbo turned on) and no CMOS didnt helped it to backdown on 4.00, I fixed that only after installed Smart Quick Boost and switch it to autotune.
Then my Ram frequency were moderated to 13,33 by default, when it must run at 16.
Maybe all trouble in Auto settings?
Cause its like it shutdowns because it reaches some values managed in settings, that force PC to shutdown before preventing damage.
 I found power protection settings in Easy Tune - 3d Power - Phase and Voltage.
But I am afraid to manage them without help.
In Voltage - CPU protection Vrin is 350mV for example.
So guys plz, can you help me with my problem, cause I didnt found working solution for me in net and forums.

[shadowsports]

You've done some testing and have performed a good amount of troubleshooting.  Is your memory on the QVL for your board?  Are the DIMMs installed in DDR3_1 and 2 (orange slots)?  3 and 4 will work but are not recommended when only 2 DIMMs are installed. 

Do you see any error information in the System Event Viewer prior to the restarts?
 
After flashing, did you...  select "Load Set Up Defaults and F10 to save and exit?
After doing so, does the default clock speed for your CPU show as 4Ghz?
Does memory show as 1333?
Now enable XMP for memory and select "Profile1"
Memory speed should now show as 1600Mhz (ensure you save this)
On restart, verify:
If you have any residual OC settings or Profiles saved, delete them and again, Load set up defaults, then F10 to save and exit.  Once again you will lose XMP so be mindful of this.

I'm using Auto settings (in BIOS) for most everything with no issues, XMP for my mem. Try testing with only the minimal number of BIOS changes needed for your system to boot.

Last:
Are you convinced all traces of OC'ing you may done previously are gone?
Consider memory as a possible cause if not on QVL
Consider rolling back or updating to an earlier / later version of GPU drivers

Any other troubleshooting steps you've tried or observations you think are relevant to the issue?

[HunkUmbrella]

Thx shadowports for your quick reply.
1.Yeah, my RAM mounted as said in manual, both of them are in orange slots.
2.Yeah, I checked the System Event Viewer nothing unusual before crash.

3.Yes, after flashing I selected load set up defaults, and it never helped, then I flashed it again to last F7, still hard resets occur.
4.Nope, after restart it goes back to 4.40, i will try to disable Intel Turbo boost in app center, it is set to auto.
but when I selected autotuning it gave me error - open driver handle failure. And after it boost CPU to run on 4.60, and you know I tried this in WarThunder and all was stable for some hours.
 After WarThunder I tried Witcher  3, I had problems with GPU drivers and OBS, reinstalled them, all was ok, but suddenly it hard reset again, maybe its badly factory overclocked?
I flashed my GPU bios from F2 to F3 and all was more or less, but for test issue, I flashed the old F2 version that I backuped earlier and it shutdown in 30 min. May it be connected?
I ll try to flash back to F3 and test more.
5.I used profile 1 and improve Ram to 1600, stability increased, but still shutdown occuried during GPU hard working.
It seem me need to prepare buy new PSU.
Ill attach screenshots later.

[shadowsports]

You think its the PSU?  How old is it?

Has decent reviews
Active PFC
80+
Plenty of punch (amps) on 12v rail

I'd try: (in this order)
Different version GPU driver
Different memory
A different PSU (unless it super old, then move this up one)

I only have AppCenter and SIV installed.  Uninstall Easy Tune.  Restart.  I have the same CPU.  It shows as 4.0 and 4.4 when Turbo Boost.  It shouldn't be idle at 4.4.  If you are seeing 4.6 something is stuck or hung up.  I'm sure your Zalman is up to the task.  4.6 on air isn't a big deal, but something is breaking.  What about airflow...  do you have a good case?     

[HunkUmbrella]

PSU - maybe, because it is PSU or GPU, well I wish it is not GPU
Yeah, I reinstalled my APP center to newer version dont know why it just didnt updated, cause I launch check for updated everytime. well now my EZ setup install freezes on 40 percent of installation and do nothing, I left it to install on whole night and still 40 percents, restarted, tried again, same thing, i will uninstall it and will try again.
Time for screenshots.
Ahh attach restrictions, ok I will upload it on Yandex disk and link you:
1.
https://yadi.sk/i/69yIr78WhkKoQ
2.
https://yadi.sk/i/vQ5y3qFohkKpo
3.
https://yadi.sk/i/7t71NSvdhkKqW
4.
https://yadi.sk/i/Y6aWcvWshkKrB
5.
https://yadi.sk/i/YFqG7-kFhkKru
6.
https://yadi.sk/i/xAK55SHIhkL5d
7.
https://yadi.sk/i/D-74_Fw0hkL7s
8.
https://yadi.sk/i/gg7rC1BghkL9K
9.
https://yadi.sk/i/ujyoJy4uhkLAr

[HunkUmbrella]

I installed all Gigabyte tools, and start recording stats during gaming, so thats the stats before shutdown occurs

LogOut_20150708.... recorded by OC_Guru2

and the second - in zip archive siv2015-07-08.csv is system information viewer log files.

I wish it could help to find the key of the problem

Is it normal that CPU and +12V are so jerky if compare with other stable Voltage lines!?

[shadowsports]

Greetings,
I still think you should try another graphics driver and if possibly different memory.  You never said how old the PSU was?

12v - The voltage fluctuations you are seeing look ok.  They are in 1/1000 volts increments.  I'd only be concerned if they went below the 12v threshold.  I agree your CPU voltage is dipping more than I'd like under load.  Temps look ok, fan speeds are linear and appear to be responding directly to the load.

I hope you are not using @BIOS to upgrade your board.  Don't risk it.  I recommend Q-Flash.     

[HunkUmbrella]

I have my PSU like two years, sry that i forgot to type that.
Yesterday I pull down CPU settings - processor graphics clock from 1300 to 1000
Active core to 40, yesterday it was stable for 2 hours, but like I said, it shutdown randomly, so I will try testing.
About @Bios, yeah first time I used Q flash to upgrade, I read about @Bios problems, but last F7 bios I dont remember why, maybe failed to upgrade from Q flash, i updates with @bios, want me to try reflash it from Qflash? Ill try
Whats about GPU bios, mine is back to F2, should I flash it to F3??

Well, i suspect PSU, cause the same happened with my GTX 690 and it died
(i dont know why, cause it spent some time I even didnt notice, cause of HD4600 graphic, I understood that my GPU is not responding or not even seen, when I wanted to check drivers in GeForce and it tell me, hey man you dont have any GPU, I was in Cyprus that time, and when i returned to Moscow and bring it to service, they said that chip was cooked so hard, I was so shocked, so now I never use motherboard HDMI, use only HDMI on GPU, to know when it going to crash.)
I dont want to loose my GTX 980 )

[shadowsports]

Still recommend you to try a different GPU driver.    

[HunkUmbrella]

My GPU driver is 353.30 by Nvidia, well why I think that is GPU drivers? Cause I update them everytime, and the problem appeared long ago, I even clean install drivers on new Win7, after I had installed gigabyte provided nvidia drivers
The most terrible things that all say, that it seems ok with my hardware and we cant seek what causes that problem.
Also I stress tested my GPU yesterday with FurMark, and it was ok, it holds up on 71C - stable and even video driver didnt crashed.
I ran Prime95 for 3 or 4 hours, CPU went more then 50 degrees but system were stable still, I will try it to run like 5 or 8 hours.
I will try to check my Ram again using Memtest86
Little question why my OC Guru 2 have max possible resolution 1600, when I got 1920 on 1080!?
I use EVGA Precision to boost my fan speed

[HunkUmbrella]

Changing and reinstalling drivers didnt helped,
Flashed vga bios to F3, using hdmi cable,
Reflashed motherboard bios from f7 to f7 using q flash, loading default options - dont solve the prob.
I noticed one strange thing, when I am ingame, PSU start to making sillent sounds, noises, and if he makes them, then pc shutdowns everytime, I tried to use different PSU-GPU cables, still it makes sounds in game, still shutdowns.
Are these noises a sign of trouble, or it is because it operate more power consumption!?

[shadowsports]

Greetings,
What happens if you take the sides of your case off and maximize airflow.  Do you have a fan you could use for this?  Running out of ideas.   

[HunkUmbrella]

I did this, my Z97 soc force is on my table in "demonstartion mode" with Gigabyte provided bracket for GPUs.
Thats what OC_Guru 2 detects before shutdown -
Date Time      GPU Clock(MHz)   Memory Clock(MHz)   GPU Voltage(V)   Power(%)   GPU Temperature('C)   FAN Speed(RPM)   
07-29-2015 08:05:12   1354      7012         1.193      62         69         2172   

07-29-2015 08:05:15   1354      7012         1.193      69         71         2167   

07-29-2015 08:05:18   1354      7012         1.193      67         71         2167   

07-29-2015 08:05:21   1354      7012         1.193      68         72         2172   

Maybe something wrong with voltage?

[dmdilks]

I did read on reviews. Some of the people are having the same problem. They are say when it hits 70c the computer crashes. If I'm reading this right "GPU Temperature('C)". Then the card is running at about 70c/158f. That is way to hot.

Yes the card does get good reviews but you might have a bad one. Now thing is some of the cards do come OC. Did you try to under clock a little just to see if help. Some of the newer cards don't need as much power as the older ones.

https://forum-en.msi.com/faq/article/power-requirements-for-graphics-cards

[shadowsports]

Agree.  Sure does look like a heat issue with the GPU.

The highest temps for GPU I reviewed in your log above was 57* C, well within safe operating range, which was why I said they looked ok.

Here you are showing us 72* C and who knows what its actually hitting before it stops logging.  This looks pretty suspicious.   

Water block may be your next investment.