Installed GTX 980 - Computer freezes after 5-30 minutes of gaming.

snaggletoof

Prominent
Sep 7, 2017
4
0
510
Hey all,

New to these forums, but I think I've gathered enough info to be useful here.

System:
**mobo has 2 DIMM slots and I have 2x8 in there right now, but it's only reading 8GB - working to resolve (either bad stick of RAM or DIMM slots are now faulty). Will update shortly.

Previously, I had an AMD Radeon 6950, but my friend gave me his old/working GTX 980 since he upgraded. The only hardware change that has occurred is this video card swap.

Immediately, I started running into issues where, when playing Overwatch, my system would seize up completely. Sometimes, everything just stopped. Other times, everything would stop & my speakers would buzz indefinitely. Only thing I could do in either scenario was hard restart each time (holding the power button).

At the time, I was running Windows 7, so the following steps have been taken since:

  • Upgraded system to Windows 10 64-bit [clean install]
    Installed latest nVidia drivers [which existed for Win 10 only]
    Set CPU fan to highest setting (why not)
    Updated BIOS to latest version
    In nVidia Control Center, 3D Settings -> Configure PhysX -> Processor: GTX 980
Each of these steps were taken independently, and none have resolved the issue.

I ran Open Hardware Monitor & took a photo of where things were for two separate freezes (couldn't screenshot b/c computer is frozen):

Freeze 1:
7YHvaoH.jpg


Freeze 2:
230gV72.jpg


Some of those temps look a little high, but maybe not enough? I'll let y'all decide.

I have also tested the video card. My monitors are 1080p, so:

baAzAew.png


None of my efforts have resolved the issue. So, curious to hear what y'all think. Happy to provide more info, if needed!
 
Solution
Your CPU cores are all at 90 to 100 degrees, so most likely your CPU is overheating. Do you have a properly mounted cooler in place? You might have bumped it loose when installing the new card. It would probably be worth cleaning and reapplying thermal paste to your CPU, and reinstalling the heatsink.
Your CPU cores are all at 90 to 100 degrees, so most likely your CPU is overheating. Do you have a properly mounted cooler in place? You might have bumped it loose when installing the new card. It would probably be worth cleaning and reapplying thermal paste to your CPU, and reinstalling the heatsink.
 
Solution

snaggletoof

Prominent
Sep 7, 2017
4
0
510


Update: I might have a DIMM slot issue. I tried the following:

  • Stick A in Slot 1 = 4GB in DxDiag
    Stick A in Slot 2 = 4GB
    Stick B in Slot 1 = 4GB
    Stick B in Slot 2 = 4GB
    Stick A & B in 1 & 2 = 8GB
    Stick A & B in 2 & 1 = 8GB
So it's only reading 4GB of each stick, regardless of which DIMM slot.



Here's an image of the inside:
OzocJRX.jpg


(A)The fan & heatsink it came with is mounted on there just fine and was recently cleaned out - perhaps a diff fan is needed? I will reapply thermal paste & report back.
(B)It might be worth noting that I had to remove the fan shroud from the GFX card so it would fit in my case. However, I don't think this is impacting anything because the GPU temp is fine, per those screencaps.
 
Your CPU temps are crazy. And with stock speeds, you should be just fine with stick cooler, my guess is is all 4 posts are not "clicked" in place":

Option 1 - Remount Heat Sink

a) Purchase a bottle of 90% or better isoprophyl alcohol and some G751 TIM. Grizzly Kryonaut is a better TIM but it's hard to find and cost 4-5 times as much, outside that, nothing is better than the Shin Etsu
https://www.walmart.com/ip/91-Isopropyl-Rubbing-Alcohol-32-fl-oz/33057974
https://pcpartpicker.com/product/mDPfrH/masscool-thermal-paste-g751

b) Clean the heat sink and CPU with the rubbing alcohol and non fraying cloth or paper towel like product

c) Apply TIM - https://archive.benchmarkreviews.com/index.php?option=com_content&task=view&id=170

d) Remount Heat Sink

Option 2 - Buy a 3rd party cooler ... you shouldn't need it but what they hey ... if you ever get an overclockable CPU, you all set.

a) Check your case specs to see what cooler heights will fit inside

b) This 3rd party cooler matches the thermal performance of the $90 top of the line products from Noctua (NH-D15) and Cryorig (R1 Ultimate) ... and it costs $37. Follow same install instructions as above

https://pcpartpicker.com/product/8YrcCJ/scythe-cpu-cooler-scmgd1000

EDIT: You see those 4 posts ate each corner of the cooler in your red box .... at least one of those I bet is not secure.

EDIT 2:

2 x DIMM, Max. 16GB, DDR3 1600/1333/1066 MHz Non-ECC, Un-buffered Memory
Due to Intel® chipset limitation, DDR3 1600 MHz and higher memory modules on XMP mode will run at the maximum transfer rate of DDR3 1600 Mhz.

Use CPUz to view your SPD settings. Should look something like this:

417x402px-LL-59cd4535_CPU-Z20SPD.png


Boot to BIOS and try the 800 Mhz JEDEC setting, save and exit BIOS

If that doesn't work, call Asus Tech Support.

Note this is unrelated to your 95 - 100C CPU problem


EDIT 3:

I am suspicious of the temps shown in Hardware monitor and you did not indicate what load was on the GFX card at the time. Run Furmark and see what happens ... if you are hitting 84C (Furmark shows temps on screen... so don't load HM) you can shut it down as the card will be throttling already.


 

Yeah, you probably bumped it loose when cleaning it or installing the card. The fan on the CPU cooler is most likely fine, and you should not need to force the fan to its highest setting to keep the CPU cool at stock clocks. That will just make your system noisier than it needs to be. Also, make sure you didn't force it to its lowest speed instead. : P

90-100C is not at all a normal temperature range for a CPU, and typically temperatures shouldn't get much over the 70s under load, so the overheating processor is most likely causing your problems.

The removal of the graphics card's shroud shouldn't likely be a problem, since that card appears to exhaust its air inside the case anyway.
 

snaggletoof

Prominent
Sep 7, 2017
4
0
510


Will try this - need to buy new thermal paste. Just tore the apartment open looking for it & no luck.

Option 2 - Buy a 3rd party cooler

This will be my last option. I'm not OC'ing so I think best bet is remounting.

EDIT: You see those 4 posts ate each corner of the cooler in your red box .... at least one of those I bet is not secure.

All 4 are secure.

Use CPUz to view your SPD settings.

Boot to BIOS and try the 800 Mhz JEDEC setting, save and exit BIOS
k2bu97h.png

This is what it was already at. Weird that it's showing that the stick is 4GB, and that it's the 4GB part number...the sticks clearly say it's 8GB on the exterior...

I am suspicious of the temps shown in Hardware monitor and you did not indicate what load was on the GFX card at the time. Run Furmark and see what happens ... if you are hitting 84C (Furmark shows temps on screen... so don't load HM) you can shut it down as the card will be throttling already.

Furmark screencap was included in OP - max of 67C - not worried about that :)

 

snaggletoof

Prominent
Sep 7, 2017
4
0
510


Re-applied paste & re-fastened the heat sink & it looks like we're good to go.

Here's what it looked like when I took it off. This is the original paste that came on the sink when I bought the processor:
M0lqvok.jpg

2nYQA5b.jpg


So, I cleaned it off w/ isopropyl alcohol:
uu4N9N6.jpg

sdamthz.jpg


And applied new paste:
TAVdHsl.jpg


And this is the new temp + max (played Overwatch for an hour for those Max figures)
GjZNqVO.png


Seems like this problem has been solved!

BONUS: I don't live near a place that sells thermal paste, BUT Amazon Prime Now sold it. Since there's a $30 minimum on that order, I had to get a bunch of booze to get the paste delivered:
PK2QDyi.png