CPU+GPU Temps Slowly Rising to 100C on minimal load

[Michael Dowden]

Hey all
I've had my most recent build for around 14 Months now and have had very little problems with it. However yesterday I noticed that when idle my CPU temp was fluctuating between 50-60C and when i opened a game(WoW) my CPU temp would slowly rise to 100+ over the course of around 5 mins. I also noticed today whilst doing another test to confirm this, that my GPU temp seemed to also skyrocket towards 90C.

Yesterday I did see that when i brought my cpu under around 15% load with the game open that only one of the tubes in my H100i was becoming hot and the other was staying cool. After scouring forums I saw that a number of people put this down to either a pump failure or blockage in the loop somewhere.

I was all prepared to replace my H100i but I don't see how that would account for my GPU temp rising so high.

System Specs:
NZXT Phantom Full Tower (Pull from sides, Push up through H100i radiator)
Asus P8Z77-V pro Mobo
i7 3770k at stock(3.5) Voltages set to auto and rarely go over 1.2v
EVGA GTX680 (Never had any temp problems before, not that WoW would)
4 NZXT case fans on pull
H100i with rad+dual 120mm fans at top of case on push


Even when my cpu reaches 100C under such low load, the air from the radiator is very cool which makes me think that there is no circulation around the cooling loop. I thought that a simple replacement would help, but then I noticed the increased GPU temp also. GPU voltage is 0.987V

Any help would be massively appreciated!

 

[Eduello]

It would indeed seem like the H100i needs to be replaced.

What test did you use to get the GPU temps to 90C? Is your GPU using the reference cooler or an aftermarket cooler like the EVGA ACX?

 

[Michael Dowden]

It would indeed seem like the H100i needs to be replaced.

What test did you use to get the GPU temps to 90C? Is your GPU using the reference cooler or an aftermarket cooler like the EVGA ACX?

I haven't made any adjustments to the GPU from out of the box. No OC or aftermarket cooling. I have the fan curve set to auto, so that kicks in after 60C.

The test I ran(and just repeated) was literally running WoW in ultra setting for around 5-10mins. That's how long it takes my CPU to reach 105 at which point I instantly closed WoW and the CPU temps dropped back to 65. At it's highest the GPU reached 75 on that last test with no real sign of slowing down.

I am running dual screen, so maybe those GPU temps are acceptable?
I'm glad to have a definitive answer on replacing my water cooler though, thank you

 

[Eduello]

Aftermarket cooling means a cooler added by the company, in this case EVGA (I know, it's an odd term).

This is a reference cooler:

This is an aftermarket cooler:

Which one do you have?

Also, what is your room temperature?

 

[Michael Dowden]

I have this one
My room temp is currently 15C

 

[Eduello]

Well those models tend to get pretty hot, but with a room temp that low it shouldn't reach 90C...

Try blowing out dust from your GPU. You can also try changing the fan curve manually (can be done with MSI Afterburner) if noise does not bother you.