99 degree celcius temps with corsair h100i GTX?

Oct 8, 2013
714
0
10,990
So I have had a few issues with my system blue screening on occassion during heavy load, but never thought much of it until now. I'm just rendering a video in sony vegas, 100% cpu usage and my system died. I've taken a look with 3 seperate software applications to see my cpu temp and they're all reading high 90s on full load.
So obviously my cooler isn't working correctly.
I'm wondering what I can do to remedy this?
One pipe is physically hot and the other cool, so it seems like the coolant is circulating, could it be a simple case that I need to dust it? Then again I just took a look at it and although there is a noticable layer of dust, it is really a thin film and I wouldn't expect such a minimal amount to cause such disastrous cooling.
Anyone have any ideas?

*edit*
Idle the cpu is sitting around 50-60 degrees. Crazy.
 
Solution
People have used all sorts of things as a TIM, tooth paste, denture cream, etc. etc.
Troubleshooting is about eliminating potentialities. Something that won't cost you anything but about 20-30 minutes is to pull the AIO out, clean it thoroughly, redo the TIM and plug the 4 pin connectors into the mobo and start monitoring fan speeds and radiator speeds doing all sorts of different tasks. How are you controlling and monitoring the AIO? Is it through the mobo or does your model have corsair link?
I don't think the radiator has died, it might not be operating at peak, but it seems to still work, as evidence by your simple test of temps on either side of radiator.
I've never had a corsair AIO, I've got several cooler masters and an...
that's a very warm idle. how old is the cooler? if it's older than 3-4 years, it should be replaced, as the coolant tends to, over time, evaporate away. if it's not old, then using the monitoring programs, you should check the stated pump speed(cpu-z is a very good program for this, so long as the pumps 4 pin connector is connected to the mobo) as compared to pump speeds in the specs. after that, it never hurts to reseat the cooler plate and the cpu and redo the TIM.
 
shoudnt both pipes be reasonably cool? the flow of coolent is what keeps the components cool, like running water.

the thermal paste applied thin and even? > try repaste
can you hear the pump increase in speed when under load? no = pump dead
are the rads warm/hot where the fans are trying to cool? do they increase in speed under load (depends on system you have)?
 


Actually that is a strong sign it isn't being circulated. If it was, both would be mildly warm at worst. I am quite sure the pump isn't working correctly. You may want to dig out the stock cooler or another backup and use that instead and see if Corsair will replace or repair it.

 


I have a i5 4670k OC'ed to 4.3mhz
The cooler now is about nearly a couple years old, with thermal paste as equally old.

Mobo- SLI krait edition
RAM - 2x8gb hyperx
GPU - GTX 1080
 


Fair point, just for more info, when I change it from performance to quiet i hear a small rumble from it, so I naturally thought it was still kicking. Perhaps not
 
So a very curious update.
Just turned off my computer, fiddled about with some wires and blew some dust away.
Rebooted and after a minute the fans have suddenly ramped up to full speed along with the pump.
Just did a stress test, but the cpu temps ramp up to 90 within the very first second, so still not working.
 
Ok, it's still got coolant, unless you've got a leak(usually happens cause the pump bearing is shot, but that's an audible noise that's very obvious).
At this point, it's either the pump isn't pumping, or the bubble gum you used as TIM is shot😛.
You don't have an absurd overclock, that takes that out. How warm is the room your computer is in?
 


Not sure I get the bubblegum TIM thing you mentioned.
I live in England, so that really takes the environment being too hot out of the equation.
But it's about 24 degrees. tad hotter than a normal room as I leave my pc mining when i go off to university for the day
 
People have used all sorts of things as a TIM, tooth paste, denture cream, etc. etc.
Troubleshooting is about eliminating potentialities. Something that won't cost you anything but about 20-30 minutes is to pull the AIO out, clean it thoroughly, redo the TIM and plug the 4 pin connectors into the mobo and start monitoring fan speeds and radiator speeds doing all sorts of different tasks. How are you controlling and monitoring the AIO? Is it through the mobo or does your model have corsair link?
I don't think the radiator has died, it might not be operating at peak, but it seems to still work, as evidence by your simple test of temps on either side of radiator.
I've never had a corsair AIO, I've got several cooler masters and an enermax. I control speeds on fans and radiators using the mobos fan controls, so if it's an issue with corsair link, I'm hopeless to help with that, other than recommend you use the mobo, set up custom fan curves, and run the radiator at 100% at all times.
 
Solution


I'm really not sure what you mean by TIM? I've never done a TIM on this, whatever that is!