Does Thermal Paste Actually Matter?

Sep 29, 2018
25
0
30
Hey guys,

The temperatures of my i7-6700k have been rising recently (I have a H100i v2) and I have been wondering whether replacing the thermal paste will help.

So, the most common question, does thermal paste matter? I've seen many highly regarded TIMs but ThermalGrizzly Kryonaut seems to be the best, any thoughts?
 
Thermal paste matters if your are running high overclocks. If you are running at stock or small overclocks, then it really does not matter. Maybe a couple degrees difference. The 6700k has not been out that long, so I am not sure if the thermal paste has gone bad. You may have an issue with your pump, thus the higher temps.

I have had issues with a Corsair AOI and the chip gradually got hotter over time. I had the cooler RMA'ed and I think that one of the tubes was clogged. The RPM on the pump in Corsair link was higher than the 3000rpm and on of the tubes was hot. After a little digging it seemed this is a problem that has happened to others before.

I would try and replace the thermal paste first, just because it is the easier solution. But it may not work, so keep that in mind.

As far as Thermal Grizzly Kryonaut, it is considered to be the best by the professional overclockers. So if you want the best, then go for it. That is what I use on my rigs.
 
Sep 29, 2018
25
0
30


Just a heads up, one of the tubes is SUPPOSED to be hot, it is the one coming out of the CPU block. Also, the reason I'm looking at reapplying is because I intent to overclock my CPU as I finally have time to fiddle with my computer etc. Thanks for the comments though :)
 
Differences in thermal pastes are usually only good for a few degrees better cooling, but if you're looking at more expensive thermal paste to solve a thermal problem, you're looking in the wrong direction.

Thermal paste that has dried or stratified should be replaced.

A good thermal pad should also get you within a few degrees of a good thermal paste, and should be a permanent solution.

Tubes near a water block should not show a significant temperature difference, provided the water is flowing through the block correctly. If the water is staying in the block long enough to show significant temperature differences before and after, I would consider there to be a flow problem.

The only tubes that should show much difference are the inlet and outlets to your radiator, and the difference they will show will be directly proportional to the ability of the radiator to transfer heat from the water.
 
Sep 29, 2018
25
0
30


I mean it's not like one tube is boiling and the other is freezing, you can just tell by holding them, one is warmer and the other is cooler than it, not specifically cold.
 
You shouldn't be able to gauge the temperature difference near a water block with your hands unless you have had them calibrated correctly, and once your water is heat saturated, I doubt you'll feel the difference at the radiator either.

Is your pump speed adjustable and running too slowly?
 
Sep 29, 2018
25
0
30


It's a H100i v2, so the pump is running at roughly 3000rpm. Considering Aseteks cheap design, the pump rotor thingy is inside the CPU block, so I think water can be expected to stay there for longer than a custom loop.
 
many AIO's temps begin gradually rising after a year or two due to contaminants, hard water deposits blocking fluid flow in pump, loss of fluid pressure due to micro leaks/evaporation...

I don't find most such AIOs to be worth more or even superior to a simple quality large conventional double heatsink/fan assembly...

 


Yeah, one of the tubes is supposed to be hot, but it was pretty hot at the radiator and it did not get hot like that before. As mdd1963 said, overtime the flow will get blocked and the hot water was sitting in the tube by the radiator for longer than it should have. Thus, it gets pretty hot.