Reapplied thermal paste, CPU still hits 80-85C under load. 35C ambient temp.

HumbleWayfarer

Distinguished
Sep 8, 2010
6
0
18,510
CPU: i7-4790K (Not overclocked)
CPU Fan: Noctua NH-U14S
Thermal Compound: Arctic MX-4
MOBO: Asus Z97-A
GPU: ASUS ROG GTX 1080 STRIX
RAM: Crucial Ballistix Sport 16GB (2x8)
OS: Windows 10 64-Bit



Ambient room temps are usually around 30-31 and can reach up to 35C on a really hot day. The room the PC is in has no access to ventilation / air conditioning and I don't have the ability to add any wall-units to cool the room. I am using a Noctua HSF that I just recently cleaned as well as Arctic MX-4 thermal paste that I reapplied a week ago. I didn't notice any change in the idle CPU temps, as they're still running 40C-50C under minor load, and 80 consistently under heavy load with 85C being the max temp a core will seem to reach at any given time.

I know there are a lot of differing methods for thermal compound application, and I aim to try a few different ones to see if it helps at all, but is there anything else that I can do here to try to bring these temps down? Someone said that going to liquid cooling wouldn't help at all, but I would think that because the ambient temp is so comparatively high that air cooling is already inefficient because I'm just circulating more hot air. Is liquid nitrogen a reasonable cooling solution if I could have it installed by someone who knows what they're doing? Would that require a lot of regular maintenance? I certainly wouldn't do that myself, but my living situation isn't changing in the near future and I need to use my PC under load. I want to do whatever I can to protect the components I have because I already just invested in a new GPU after my last one starting shorting the PSU overnight.
 
Solution
That's a quick no, to liquid nitrogen cooling as a long term solution. When it is used, it is poured manually into open air pots where it boils away. The motherboard has to be carefully prepared before hand to prevent damage from the condensation that occurs with the cold temperatures.

How one applies their thermal paste is a matter of debate and if you're already applying sufficient amounts, is unlikely to do anything but waste both your time and thermal paste. Application results are reasonably well understood. Too little thermal interface material and you won't get good heat transfer, too much and you can end up with a mess, but it doesn't inhibit the transfer provided the mating surfaces are correctly joined. The excess thermal...

Dunlop0078

Titan
Ambassador
I would think that cooler would do a bit better than that.

Liquid nitrogen is not reasonable at all, it is used almost exclusively by extreme overclockers looking to break benchmark records or the like. You would need to keep buying LN2 to use it constantly and condensation is a real concern over time.

What is VCORE voltage under load? A program like HWInfo will tell you.
 
That's a quick no, to liquid nitrogen cooling as a long term solution. When it is used, it is poured manually into open air pots where it boils away. The motherboard has to be carefully prepared before hand to prevent damage from the condensation that occurs with the cold temperatures.

How one applies their thermal paste is a matter of debate and if you're already applying sufficient amounts, is unlikely to do anything but waste both your time and thermal paste. Application results are reasonably well understood. Too little thermal interface material and you won't get good heat transfer, too much and you can end up with a mess, but it doesn't inhibit the transfer provided the mating surfaces are correctly joined. The excess thermal interface material will just squeeze out. This is a bigger deal if using electrically conductive compound.

With this in mind, make sure you used enough of your thermal paste, and make sure the heat sink is mounted as best you can. Sometimes it means laying the chassis on it's side so the heat sink can free stand and be held level to get a consistent torque on each hold-down screw.

Finally, air and water aren't all that different in the end. Yes, water can absorb more heat, but only up to a point. Once you heat saturate your water loop, it's not going to work any better than your air cooler as the water is still being cooled by an air cooler, and it will also retain the heat longer. Water tends to give impressive results when there is enough water in the system to absorb the thermal load and return to an idle state before becoming saturated. The more water in the system, the longer it can go before it saturates.

You have to expect that the water in your loop is going to start at your already high ambient temperature after being in the environment for enough time, so your baseline starting temperature with water is going to end up being the same as your air cooler. The biggest difference is, your air cooler doesn't have much capacity for storing and transport of the heat therefore you need to ensure good heat transfer away from your air cooler. This requires your chassis to have good ventilation, and there's zero substitute for that. If you're trapping the hot exhaust from the heat sink on your CPU or GPU inside your case, you're going to see one or both heat sinks perform badly.

Are you running your fan(s) at their highest rated speed?

Are you using the low noise adapter Noctua provides?

You can add a second fan to your NH-U14S which may improve thermals slightly.

You can also try running the system without the side panel installed to see if thermals improve. This would show an airflow problem.
 
Solution