What the best thermal paste in 2019?

[Athlonite]

@lynx29 yeah they're not supposed to be to bad I remember watchin Linus do a video on it a while ago now

 

[delshay]

@lynx29 yeah they're not supposed to be to bad I remember watchin Linus do a video on it a while ago now

I'm keeping an eye on "Hitachi TC-HM03" thermal pad which is used in Radeon 7 graphic card according to reveiws. I'v used Innovation Thermal pad, but idle temperature is too high, so it could not stay on my CPU/GPU die, but it is used elsewhere inside my R9 Nano.

 

[Susquehannock]

White toothpaste is not all that silly when you consider it's comprised mainly of Aluminum oxide (alumina, aloxide, etc). Same component in many ceramic based TIMs. Makes an excellent abrasive too.

 

[lexluthermiester]

White toothpaste is not all that silly

Except that it's been tested and shown to be worse than if you just used nothing and had bare metal to metal contact.

 

[Susquehannock]

Except that it's been tested and shown to be worse than if you just used nothing and had bare metal to metal contact.

My tests were not that way. Once the fluids evaporated after a few days it was still a bit better than just slapping the sink on there. Then again I tried toothpaste on unmodified heatsink. Probably would be different if lapped flat to a mirror finish. I also tried some 10,000 grit alumina polishing compound. Seemed to work better than the standard white paste. Anyway ... was just making a comparison and happy they don't test with bologna anymore. Glad I bought the big tube of Kryonaut.

 

[peche]

shite came a bit late to the party, i do love thermal interface material threads!

Best paste for bla bla bla.... well my friends, there is no best, better or worse, is all matter of application method, what does this mean?

Application method, surface clean-out, also mounting presure and cooler might affect results, as well case airflow and ambiance temperature, so taken this out, performance and results may vary according the choice of TIM, i truly state that i love Arctic mx 4 cause it brings great performance and you could find it pretty easy out there, also it spreads easy, might be to hard for some and to easy for others, but according results this babe sits around the best sellers and best performers!

Regards,

 

[lexluthermiester]

My tests were not that way. Once the fluids evaporated after a few days it was still a bit better than just slapping the sink on there. Then again I tried toothpaste on unmodified heatsink. Probably would be different if lapped flat to a mirror finish. I also tried some 10,000 grit alumina polishing compound. Seemed to work better than the standard white paste.

Fair enough. It would be interesting to see more in-depth testing like that.

Glad I bought the big tube of Kryonaut.

It's good stuff.

 

[Susquehannock]

shite came a bit late to the party, i do love thermal interface material threads!

Best paste for bla bla bla.... well my friends, there is no best, better or worse, is all matter of application method, what does this mean?

Application method, surface clean-out, also mounting presure and cooler might affect results, as well case airflow and ambiance temperature, so taken this out, performance and results may vary according the choice of TIM, i truly state that i love Arctic mx 4 cause it brings great performance and you could find it pretty easy out there, also it spreads easy, might be to hard for some and to easy for others, but according results this babe sits around the best sellers and best performers!

Regards,

This is very true. Far too many variables for any absolutions here. Not to go off on a tangent, surface finish can be a critical factor too. Being a career machinist I am bit of a geek about those kind of things. Even within the Dell T3500 proprietary heatsinks I have seen two variations in how the contact surface is machined. One is far more smooth and flat than the other. CPU heatspreaders? They are rarely flat either, if ever.

 

[peche]

This is very true. Far too many variables for any absolutions here. Not to go off on a tangent, surface finish can be a critical factor too. Being a career machinist I am bit of a geek about those kind of things. Even within the Dell T3500 proprietary heatsinks I have seen two variations in how the contact surface is machined. One is far more smooth and flat than the other. CPU heatspreaders? They are rarely flat either, if ever.

totally true, surface is where the reason for Tim starts, also monthing presure might affect, there is no rule for mounting presure, since you cant measure it, but is important to try, research and also do this, exchange ideas...

Cleaning must have a place here, traces or leftovers of old tim could affect results.

Regards,

 

[vMax65]

Always used the good old MX-4 but just changed to the new Cooler Master MasterGel Maker Thermal Paste 'Ultra High Performance, Nano Diamond Particles....
I found it great and did better than the MX-4 but most importantly it was so easy to spread...Not bad at all..

 

[hat]

I would be wary of diamond pastes. I still remember the whole IC DIamond debacle here at TPU... that was fun.

 

[vMax65]

I would be wary of diamond pastes. I still remember the whole IC DIamond debacle here at TPU... that was fun.

Thanks for the heads up...I'll look into it.

 

[hat]

If I could go back and edit my post on the first page, I would. It seems TG Kryonaut is also a diamond paste, so I wouldn't recommend it... not without specifically mentioning that it is a diamond paste and can scuff up surfaces. As for me, I'll stick with MX-4.

 

[eidairaman1]

If I could go back and edit my post on the first page, I would. It seems TG Kryonaut is also a diamond paste, so I wouldn't recommend it... not without specifically mentioning that it is a diamond paste and can scuff up surfaces. As for me, I'll stick with MX-4.

Conductonaut I'd be wary of for sure

 

[hat]

Conductonaut I'd be wary of for sure

That's a liquid metal paste. I wouldn't even mention it, unless I knew the person I was talking to was quite experienced, or delidding a CPU (in which case, I'd hope they'd be experienced anyway).

 

[peche]

I still remember the whole IC DIamond debacle here at TPU... that was fun.

stinky fish memories, cant believe i came late to that party too. lool

 

[stoggs1]

Thanks for all the replies, I appreciate the input and I learned some new and interesting things about TIM.

I ordered some noctua nt h1.

 

[John Naylor]

It's hard to claim "the best" from a single test or even a test with limited samples. The benchmark test for product performance and application methods was the 80-way test done by benchmark reviews. If anyone every finds a link to that and the application article archived anywhere, the link would be most appreciated. The ones mentioned are mostly in the top 10 but unless all the others are in it and performance among tests is not consistent, it's hard to say anything with certainty ... that's why I used the 80-way test as my benchmark.

Another facet that must be considered however is "working time". IC Diamonds products for example are a bit thick. You can make it more workable for longer by putting a cup of water in the microwave a bit and, after taking it out, give the tube a warm bath ... don't take too long tho as it will get to room temps rather quickly. Some even warm up their CPU and heat sink with a hair dryer. Deinitely not for use with GFX cards ... no matter what you do, not going to keep it workable on 16 VRMs, 8 memory chips and the GPU ... and then flipping it over to do the backplate side.

 

[danbert2000]

Thanks for all the replies, I appreciate the input and I learned some new and interesting things about TIM.

I ordered some noctua nt h1.

Compared to Kryonaut, NT-H1 is cheaper and easier to use. Kryonaut is like a thick paste. NT-H1 is more viscous and forgiving. I saw next to no temperature change going from NT-H1 to Kryonaut, and it cost way too much.

Next time I'm probably going to try NT-H2. Sticking with Noctua is my plan, they make good everything.

 

[RealNeil]

Thermal Grizzly Kryonaut is what I use most of the time.

 

[dirtyferret]

Compared to Kryonaut, NT-H1 is cheaper and easier to use. Kryonaut is like a thick paste. NT-H1 is more viscous and forgiving. I saw next to no temperature change going from NT-H1 to Kryonaut, and it cost way too much.

Next time I'm probably going to try NT-H2. Sticking with Noctua is my plan, they make good everything.

NT-H2 review below, plus it's in French so it got the stinky cheese award making it fantastique! (or bad, my french sucks)


https://www.hardwarecooking.fr/test-noctua-nt-h2-et-lingettes-na-scw1/

 

[Susquehannock]

If I could go back and edit my post on the first page, I would. It seems TG Kryonaut is also a diamond paste, so I wouldn't recommend it... not without specifically mentioning that it is a diamond paste and can scuff up surfaces. As for me, I'll stick with MX-4.

TG Kryonaut is diamond based? Worried about scuffing surfaces? Better not use any aluminum oxide (a common abrasive) products either as it is just below Diamond on the Mohs hardness scale. Which leaves out virtually all the ceramic based TIMs.

 

[Bronan]

@MrGenius
I do not really agree that it does make such a big difference however all 3 are pretty the best you can buy and they all are expenssive.
I preffer still grizzly however i am no longer opening up my gpu's to put better paste in them ... especially because they made it much harder to open up the cards.
And you also might damage any vapor chamber if is present, which many use these days (if your not cautious enough)

 

[Vario]

NT-H2 review below, plus it's in French so it got the stinky cheese award making it fantastique! (or bad, my french sucks)


https://www.hardwarecooking.fr/test-noctua-nt-h2-et-lingettes-na-scw1/

Oh boy that 0.5 to 1 degree delta between the pastes, thats totally not in the realm of error is it? its CLEARLY better LOL /s

 

[Wavetrex]

All tests are pointless the minute after applying paste.
Apply mayo or toothpaste, let the computer run for at least ONE MONTH, and then test for temps !

That's what makes the difference between a good compound and a not so good one (or something really bad), how long it keeps it's characteristics after time passes, humidity and oxygen make their way into the substance and so on.

Another example:
Put french fry oil inside a cooling fan - Works great. Start the computer the next day... it's noisy. Let the computer sit for 1 month - The fan doesn't even spin anymore, it's completely locked solid.

~~~
Edit: I vote for TG Kryonaut.

Applied it on my pretty hot 6800K. It has the same decent and good temp delta today as it had 6 months ago when I applied it. No degradation whatsoever.