Best thermal paste for msi laptop

Well it's out there any thermal paste that gives stable temps for long time?

I have try kryonaut many times and noctua h2.
At first days the temps it's awesome.
Like 75-80c max at full clock on stress tests
Then after some days the paste get dry? The temps raise about +10c and the difference of cores too.

 

You can try Gelid GC Extreme. I personally am using Kryonaut on my GS66 without any issues, but it seems your laptop has a really pronounced pump-out effect.

 

Yea it has that 3 screw heatsink on cpu and really have not good fit on chip.i have see at right side of cpu give more pressure

 

I ordered gelid gc extreme. Kryonaut and ic graphite thermal pad.
I will try all. But for gpu kryonaut it's the best
The temps it's good but I dont have the temps that have after repaste

 

If you have an uneven heatsink with already low mounting pressure, something like ICDiamond might help since its thicker. I am not sure the IC Graphite thermal pad will help at all.

Kyonaut is great. If you properly treat your heatsink and get it even, it should give you among the best possible performance.

 

Well I did a test at last repaste.
I paste the cpu and I placed the heatsink and tighten the 3 screw at cpu 5 6 7
When I removed again the heatsink and I saw that the cpu chip had a lot of paste at left side and at right had really less. The pressure of the right side on chip was high and spread out the paste.or at left side the pressure was very low

 

Thermalright TFX is the best paste for laptops. As thick as Phobya Nanogrease Extreme and better than NE also. Tested on a 10900k to be 2C better than Kryonaut after it cures for several days.

 

How thick is it? It seems to have quite a decent thermal conductivity, I want to try it. I might try it on my desktop as well once my new cooler arrives. The Wraith Stealth sucks.

 

I don't know how to measure thickness but it feels as thick as my Arctic Ceramique from 2005(?). I still have a giant tube of that stuff. Just checked it just now. The stuff before Arctic Silver 3 took its place...actually I have NO idea how old my Ceramique is! It might be OLDER than that...

 

U Thing It's bad idea guys to not full tight the right screw (no7) at cpu?I see there has higher pressure than left.

 

Try it and find out. It won't harm anything. Just unplug the battery and ac before doing any of that.

 

Guys guys guys. Tell me that,the ic graphite its stable at use and not raise the temps after use like thermal paste.
I am supricely from the results.
I was scared that I will hit 90c with that think
I upload 2 pics.
One with noctua h2 after 20 days use.( exchalty same results with kryonaut)
And one with ic graphite after 2 hrs of use.
Tell me that I have fixed my problem!

1st pic it's the ic graphite
(Dont looking the 20watt,its the bios tweaks)

 

IC graphite pads work well if the heatsink has a decent fit.
The temps you get are the temps you keep. The pads won't dry out or pump out.
With thermal paste or LM, if you get a proper heatsink fit and good pressure and insulation, your temps will be better.
If there are any imbalances or imperfections, the paste will dry or pump out and the temps get worse. IC pads don't care about that.

 

Just wanted to check. The first pic is IC Graphite and the second pic is Noctua H2?

If that's accurate then those temps and core uniformity are exactly what I'm looking for. I used Kryonaut on my GS66 and I notice an upward of 10C difference between cores. Here are my thermal benchmarks below. I'll probably buy the IC graphite pads if you can confirm my question above.

https://www.reddit.com/r/MSILaptops/comments/h0la7g/comprehensive_thermal_benchmarks_of_gs66/

 

Yea the first pic it's with ic graphite pad.

With fresh noctua or kryonaut I have max 78c at thit test and difference cores about 6c.
But every day the temps get raise till reach close to 90c and difference cores about 10c.
And I have try all the repaste method and carefully tight the screws.
Ic graphite for me till now the second day it's the best.
U have about +3c 4c compare to fresh kryonaut but if its stable it's the best. Coz kryonaut for me give +10c to some cores after some days and reach 90c.

*I am using kryonaut at gpu not ic graphite.
For gpu its stable for long time.

 

So my laptop doesn't suffer from increasing temps like yours, but at max load I do observe a 10C delta. I'm hoping that since the IC graphite pad is very efficient at transferring heat in the x-y plane that the temperatures will average out, so that the max core temps end up getting lowered.

I'm getting these pads on Monday and will be posting new thermal benchmarks for the GS66 soon. Hopefully this gives more uniform core temps and also lower max temps as a result.

 

Here is my 3 test with grizzly.
The first pic it's with fresh repaste (the first benchmarks)
2 pic after 1 day
3 pic after 10days+
Exchalty same with noctua h2

 

I ended up testing IC Graphite Thermal Pad on the CPU in my S66. My temps ended up being worse despite several readjustments and compressing the thermal pads on the VRMs. There just isn't enough pressure to properly compress the graphite pad on my laptop unfortunately. Oh well.

 

What u mean isn't enought to pressure compress the pad ? Isn't good pressure on ic graphite pad?
Btw I am using k5 pro on vrms

 

Sorry, my text got messed up a bit in the previous comment. What I meant was my heatsink doesn't seem to apply enough pressure on the parts of the pad nearest to the VRMs. As a result, the conductivity ended up being poorer despite several readjustments.

 

IC graphite isnt good because its too thick. As such it needs high mounting pressure which is not present in laptops.
Thermal Grizzly Carbonaut is better simply because its thinner.

The best thermal interfaces i've used so far are Conductonaut for the cpu and Carbonaut for the gpu. This is provided you dont have problems with uneven heatsinks. I've lapped mine to a mirror finish so its as even as can be which is especially important if you plan on using Carbonaut.

 

I thing carbonaut its 2mm thick.and ic graphite its 1mm. I was ready to try carbonaut but I saw that it's too thick.
And many reviews report +3-4c higher temps compare to graphite.
Dont know but this ic graphite I have try at ge75 it's the best for me. Its give u about 3c high temps compare to very fresh kryonaut but never go higher like paste does
Till now and after a week about ,its stable like the first minutes of installed

Edit: no maybe u are right carbonaut I thing its 0,2mm.
But many reviews have report higher temps than ic.
About 3-4 I thing carbonaut need more pressure for heatsink for works better

 

Carbonaut is 0.2mm thick and its very soft so it can be compressed to be even thinner than that. Carbonaut is like a very soft cloth while IC graphite is quite rigid.

Softness is very important as it allows the pad to conform to the unevenness of the surfaces better.

Carbonaut is single use though. Once it's compressed it will tear when you try to remove it.

On the gpu side carbonaut performs better than kryonaut and is on par with Liquid metal for me but doesn't have longevity issues which is why I switched over to it.

 

I was actually interested in using Conductonaut on the CPU in my GS66, but I wasn't sure whether it was worth the trouble. As noted in my previous comments I have a pretty wide spread in max core temp, upwards of 10C. It also a tripod mount for the heatsink on the CPU.

Would it be advisable for me to use liquid metal in my case? I'm mainly looking at using liquid metal for lower fan speeds since if it works right it should give me at least~10C reductions in CPU temps. Currently, with coolerboost on my laptop can handle all core boost at 4.2ghz(65-67 package power draw) with temps below 90C fine using just Kryonaut.

 

As long as you watch some good LM application guides theres no harm in trying.

 

Just out of curiosity how much of an improvement did you find in CPU thermals when you used liquid metal when compared to a high quality non conductive paste like Kryonaut?

Also do you have any recommendations for low density foam with high thermal thresholds that I can buy online? I'm looking to put up a foam dam as recommended in some previous posts other users have made before.

 

I just used cheap dishwashing sponge. Works fine. Keep in mind only the cpu die gets really hot. The surrounding is much cooler.

The important thing is to use very soft foam so that it doesn't interfere with the already weak tripod mount.

On the cpu I got down to 69°C at max fans compared to 92-94 stock. With kryonaut I was getting mid 80s.

Then I bypassed the cpu power limit and kryo started hitting the low 90s so I switched to LM and it dropped to the mid 70s

 

This is really the important thing.
Not that someone is 3C higher than Kryonaut, but that they will be at that 3C forever and it will never change.
So they can actually use their computer and not have to care about worse temps (unless ambients get worse by same delta).

I have alot of graphite pads. They are reliable long term. Just not the best for temps.
Never tested carbonaut. Never will either, not at those prices if you can't re-use it (it was advertised as reusable by Der8auer, remember?).

 

Yeah they cost a pretty penny. I've gone through 3 of them and am on my 4th one and I likely won't be repasting for many years to come.

The thing is, if you want to reuse them you can, but if you actually installed them in a way that they perform perfectly, they will be squeezed so thin that they will tear immediately upon removal of the heatsink.

Even if by some miracle it doesn't tear, it will not go back to its original thickness, making it less effective if you don't place it back in exactly they way it was originally placed. The accumulation of dust and other small particles on the surface will also reduce its effectiveness.

 

Hey, thanks again. I ended up repasting with liquid metal on the CPU in my GS66. I decided to use Thermalright liquid metal Silver King instead of Conductonaut, since the latter would take a week before it would arrive here. Mr.Fox also said the Thermalright liquid metal had the same performance as Conductonaut so no real losses by using it instead.

http://forum.notebookreview.com/threads/liquid-metal-and-33-tape.829130/#post-10985882

https://www.amazon.com/Thermal-Perf...ermalright+liquid+metal&qid=1592784282&sr=8-3

Also I tried the foam dam, but it ended up giving really high temps because of the metal shroud that surrounds the die in the 10th gen CPUs. This shroud is also present in the Asus Zephyrus S15, which includes liquid metal out of the factory, so the shroud is an Intel addition I think? I'm guessing they put it there to prevent thermal paste/liquid metal runoff. I instead opted to create a barrier by putting a thermal paste boundary on top of the metal shroud. All of these details are mentioned in my Reddit post below. Here is my post with benchmarks.

https://www.reddit.com/r/MSILaptops/comments/hdhdca/gs66_liquid_metal_versus_kryonaut_thermal/

Overall I noticed an 8-10C improvement in max core temps by going from kryonaut to Thermalright liquid metal. Core differentials also improved from 10C to 5C in between cores.

 

Nice, that sounds about right.

Yes, instead of a foam dam you could use paste, it'll just be messier to clean up the next time you need to reapply liquid metal.

 

Old school Arctic Ceramique 1 is a very nice dam if you have it around. It's very thick, apparently doesn't dry out and won't go anywhere once it's applied.

 

This.

 

Thanks for the recommendations, you guys. I'll take a look at using Ceramique the next time I need to repaste the liquid metal.

I also got a bit curious last night and decided to repaste the GPU as well with Thermalright Liquid Metal Silver King. In my case, there weren't any observable differences when compared to Kryonaut. Max temperatures were 62C at full load drawing 80 watts. It could be simply that the heatsink provides high enough pressure on the GPU so that the difference between Kryonaut and Liquid Metal is negligible or there just simply isn't a large enough thermal load to see its benefits. The stress tests were run for 15 minutes with Aida64+Unigine Heaven benchmark running simultaneously.

https://www.reddit.com/r/GamingLaptops/comments/he0ddb/gs66_liquid_metal_vs_kryonaut_on_gpu/

 

Yes there's not much to gain from putting LM on a Max-q gpu. There's just not enough thermal density.

 

GPU loves thicker paste specially in MSI laptops, Or you can lap your cooler and use a carbonaut like @seanwee did.

 

lm only ,,works fine ,,,tested year or more ,,cpu/gpu \ 3m tape+ silicone
gt75 7re ,,all paste die fast on 100w config

 

Speaking of thicker pastes, I just brought a tube of new cooler master Mastergel maker to repaste my zephyrus G. The one with the flat head design.

The tube is horrendous. The spread is uneven and there's a lot of wasted paste just crusting around the wide nozzle. It's a waste more paste design imo.

Paste wise it's much better than before. It's more viscous than the old version and surprisingly it outperformed kryonaut by a fair margin. On the zephyrus I used to get 69°C on the gpu before and now it runs at just 65°C. (zephyrus G is running a 1660ti Max-q).

Thanks for the gelid GP extreme recommendation btw @Papusan. It's cheaper and better than thermal grizzly minus 8 pads.

 

I might try lapping my heatsink on my GS66. Did you notice any drastic improvements in temps by doing so, and what was your process for lapping the heatsink. I was going to just follow the one from Overclockers club.

 

Appreciable but not drastic. I've tested it before on an old laptop.

Here's the post

 

I just lapped the heatsink on the GS66, I got about the same results as you did ~4C decrease and temps, but most importantly the differences between core temps are also reduced, 2-3C difference between the highest and lowest max core temps.

 

guys have any of you tried this thermal compound?
EC360® DIAMOND 11W/mK Thermal paste and EC360® RUBY 13,4W / mK Thermal paste., they talk about it well around and I wanted further confirmation.