Best Thermal Paste for CPUs 2024: 90 Pastes Tested and Ranked

Toggle sidebar Toggle sidebar
rubix_1011

rubix_1011

Contributing Writer
Moderator
Jan 12, 2007
16,707
660
60,190
Yep, I agree, but the thinking being GPU thermal paste updates are a bit more specific and not necessarily as user-friendly as CPU cooling applications, so we will likely be focusing on them as a standalone article with some reference back here. But yes, I would like to get something like that in the near future.
 
  • Like
Reactions: -Fran- and TJ Hooker
G

Glock24

Distinguished
Sep 5, 2014
91
13
18,635
Wow, how's things have improved! I've used MX-4 for like the last decade. Good to know there's something better and not overly expensive.
 
CompuTronix

CompuTronix

Intel Master
Moderator
Nov 13, 2006
6,374
279
38,240
As usual, great review, Garrott, and very thorough. I can appreciate how tedious and time consuming that work can be, as well as the research, patience and attention to detail that's required. I posted a link to your review in the Intel CPU Temperature Guide 2021.

CT :sol:
 
B

BogdanH

Reputable
Sep 21, 2020
618
229
5,840
I appreciate the work that has been put into this test, where one can see comparison for most known products.
It's interesting to see, that between (popular) similar priced thermal pastes , there's quite a difference in performance. Here I would say, the difference of 2°C (and bigger) is important enough to influence our decision.
What surprised me is, performance difference between metal and "normal" pastes is quite small -considering big "listed" conductivity differences. And that shows how important such tests are.
 
  • Like
Reactions: bolweval
S

sstanic

Distinguished
Aug 6, 2016
66
28
18,560
Nice comparison, thanks. Hovewer I would like to see Noctua's H2 paste from 2019 in the comparison, instead of the H1 from 2007. Hopefully that can still be added?
 
  • Like
Reactions: Nestea_Zen
R

Rdslw

Honorable
Aug 1, 2017
1,900
139
13,340
I have a comment, NT-H1 was slightly shorter lived paste than others, at least in my testing in laptops mx4's laptop started to show thermal issues (system throttles, palm rest is getting hot) similar to default paste after ~1.5y whole MX holds without noticeable changes for at least 3years now.
It was possibly old NT-H1, I am not sure when it was changed or what I had.
I de-dusted this 3 times in the meantime, so it's not a dust clog effect, and fresh paste made it again into a comfortable palm rest.
 
K

kilthas_th

Distinguished
Mar 31, 2010
40
0
18,530
I am sure there are other examples (and not trying to fanboi), but I was looking particularly at the results for Innovation Cooling's Diamond paste. In the 2017 charts, it consistently outperformed Artic Silver 5 and Artic MX-4, as well as Prolimatech PK-3, in the temperature delta results. However, it now finishes consistently behind PK-3, and trades blows with the other two. I realize these results are still all relatively close and potentially within some margin of error, but I'm curious if you investigate results changes like this. For example, do you reach out to vendors to see if there have been formulation changes when a product jumps up or down the charts, and/or re-test for confirmation, etc?
 
Johnpombrio

Johnpombrio

Distinguished
Nov 20, 2006
248
68
18,770
Thanks! It is FINALLY time to retire my giant tube of Arctic 5 and I gave up using liquid metal after only one application as it is STRANGE stuff to use. Got a tube of Pro PK3 coming.

BTW, the BEST way to remove paste type of thermal paste is Goof Off followed by isopropyl alcohol. Goof Off works amazingly well to remove even the most massive paste builds and no matter how long it has been on the die or if it has spread to the edges of the CPU. Goof Off, paper towel, alcohol, paper towel, microfiber wipedown, apply new paste. Clean off paste applicator with Goof Off as well. AND you can FINALLY buy 99% isopropyl alcohol again (I got 3 quarts of the stuff from Amazon the other day, heh).
 
B

brian532

Distinguished
Sep 12, 2007
27
9
18,535
An additional measurement that might be most useful is how a paste holds its thermal properties over time. I would rather know which paste can (or can't) maintain its cooling ability over 5+ years of use vs which is 2 degrees better when new. Maybe all pastes work well over the years or maybe several have a severe performance drop in just a year or two. They all seem to perform very well when new but what about in 5 years?
 
rubix_1011

rubix_1011

Contributing Writer
Moderator
Jan 12, 2007
16,707
660
60,190
brian532 said:
An additional measurement that might be most useful is how a paste holds its thermal properties over time. I would rather know which paste can (or can't) maintain its cooling ability over 5+ years of use vs which is 2 degrees better when new. Maybe all pastes work well over the years or maybe several have a severe performance drop in just a year or two. They all seem to perform very well when new but what about in 5 years?
Click to expand...

We'd be updating this article in about 2520 if we needed to accumulate 5 year duration stats for each compound. I know this is a bit joking, but also difficult to also estimate.

This is also lends to another statistic which still falls within a window of how often a large percentage of individual re-builds a normal person might do within a 5 year period.
 
B

BogdanH

Reputable
Sep 21, 2020
618
229
5,840
Something makes me wonder in this test.. for example: When Noctua is used on AIO it's on 6th place, but on air cooler it's on 7th place (both high tension). If we test paste only, that shouldn't happen, because paste has the same conductivity on both cooling solutions. Or, why is i.e. Kingpin KPx so much better on air cooler than on AIO?
What I'm trying to say is, cooler model/type shouldn't influence quality of thermal paste. Or am I missing something?

brian532 said:
They all seem to perform very well when new but what about in 5 years?
Click to expand...
I don't think that's a problem if we're using quality paste of known brands. I'm quite sure that in 5 years, we loose more cooling performance because of dust that accumulates (glues) on air cooler surface.
Saying that, if PC has been used for 5+ years, good cleaning is recommended anyway. And I don't mean only a bit of air blowing inside PC. Cooler should be removed, washed and new paste applied -if we're really concerned about cooling performance.
 
rubix_1011

rubix_1011

Contributing Writer
Moderator
Jan 12, 2007
16,707
660
60,190
BogdanH said:
Something makes me wonder in this test.. for example: When Noctua is used on AIO it's on 6th place, but on air cooler it's on 7th place (both high tension). If we test paste only, that shouldn't happen, because paste has the same conductivity on both cooling solutions. Or, why is i.e. Kingpin KPx so much better on air cooler than on AIO?
What I'm trying to say is, cooler model/type shouldn't influence quality of thermal paste. Or am I missing something?
Click to expand...

Understandable question, but the real item here for discussion is that there are very, very slim margins of difference which are based on load averages, but still well within a very small window. We're talking that the difference between '4th place and 11th place' might be less than 1C of overall change.

We're looking changes in decimal points between many of these, which is the primary point of the article and the findings - many compounds perform quite similarly when tested under the same testing conditions, or to put it better yet, do not give a significant advantage over a competitor brand.

The primary differences are only seen between liquid metals and traditional pastes where the actual thermal medium is very different with significant chemical makeup and showing a decided change, although still within a few degrees Celsius overall.

TL DR:

The 'place' of a compound listed is far less significant and only meant to order based on the averages readings done during testing and the averages of all core temps across all like tests for a chart. We could test these same pastes 1000 times over and the order might change slightly each time given slightly different system logging, however, each should be within the same incredibly close grouping in terms of overall average temp.

The takeaway is that choosing any of these pastes for usage will ultimately come down to a specific choice for each person, but price seems to be the first logical decision, then followed by performance. Given that nearly the same performance can be had at a range of different prices, the choice can easily be one for budget and brand loyalty.
 
  • Like
Reactions: BogdanH
_CodeZ1LLa_

_CodeZ1LLa_

Dec 26, 2020
10
10
15
I using the Thermalright TF-8 paste for all my reviews and testing. It proved to be the best so far, but I didn't try KPX paste yet (about to order it).
This paste has 13.5W/mK, it's very soft and easy to apply. It dropped CPU block (TechN) C/W from 237 with Noctua to 215 with 5900X.
It really MUST be added to this review, because I can bet it will give a really hard time to all competitors.
P.S.: They also make outstanding thermal pads. I used FujiPoly on all my GPUs, tried Thermalright pads this time with 3080 and they perform exactly the same for the fraction of the price.
 
Flyfisherman

Flyfisherman

Distinguished
May 29, 2012
1,136
8
20,365
This article has some serious problem, it has some serious questions unanswered.

First of all: how did you applied the cooling paste?
Do You really know how a cooling paste works?
It seems to me that you don't.

No worries I will explain.

For instance, it is not so much about the type as it is much more important of how.

Don't ever think applying a thick cooling paste will solve a temperature problem! Actually it will worsen this!
The reason is:
- The only sole purpose of a cooling paste is to even out the microscopic uneven and un-leveled surfaces between the component and its cooler.
NOT to create a layer of paste between the parts.

Metal to metal is of course the best cooling and if one put to much paste in between, well that is not an optimal way of how to do it.
Cooling paste by its nature has worse temperature transfer than a clean metal to metal.
That is the reason for my advice below:

Just a small, very thin layer of the cooling-paste over the surface - Not more.
That is enough to fill the gaps in the microstuctrure and every un-even surface.
So, how to do this: Just a small dot in the center of the CPU or preferably use a a credit-card (or similar) and with the same amount of small application, just smoother it out - remember it has to be thin, very thin almost unseen.

Otherwise the cooling paste will serve as a another poor heat spot.

  • Also it will take approx 15 to 20 temperature cycles before the cooling paste gets to its optimal performance, meaning for most it will then drop temperatures from 2-5C.
  • Why? Because it is a paste and it will take some time to evaporate its solvent to become a dry heat transfer, its fumes has to go away.
Temperature cycles in this regards meaning you run the system at full power and then shut i off and wait to cool down, which will be made by natural usage, no need for excess.
  • But for your so called "test" it has to be done for every single paste - No way around.
  • Then perhaps You will find something different on the temperature readings for every paste you tested.
But Your article here does not mention anything near of this or how you applied the different paste for the test nor did you take up that the solvent in its paste has to go before you can actually take a reding of the temperature and therefore I strongly suggest that this article should be seeing as a subjective test and Not qualified as a serious objective test.

Best regards from Sweden.
Ps. I am sorry for my poor English and grammar.
 
Last edited:
  • Like
Reactions: ckelih
H

hotaru251

Splendid
Oct 30, 2014
1,611
1,207
21,740
the "best premium"

Price per gram: $3.60

yet shows no downside.

"runner up best premium"
Price per gram: $9.99

downside cost per gram....understandable.

"best premium alternative"
Price per gram: $2.57

con moderate cost....


#4 also has same con yet near same as #3's cost....




Can you explain HOW these are decided on?

if #3 and #4 have cons of moderate cost per gram...yet cost LESS than #1 who doesnt have a con...that makes no possible sense.

#1 either deserves the con of cost per gram as those others or those others shouldnt have a con.
 
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom