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

[LukeHK45]

Any reviews on Honeywell 7950 phase change thermal pad? I've read individuals giving it high praise but have not seen any performance comparisons or anyone really talking about this stuff.

There's also the thermal compound version called Honeywell PTM 7950SP, from what I've heard about it, it seems to one of the very best if not the best thermal compound for laptops, for desktops I assume it would be among the top performers as well. Would be interesting to see it in the benchmarks to see how it performs in desktop applications. I will probably buy it just to check it out anyway, I've been testing Chinese OEM thermal compounds and have been pretty impressed so far, I'm sure Honeywell PTM 7950 will be pretty impressive too.

 

[sonichedgehog360]

As a note to those reading this review and naturally concluding "man, Kryonaut is da bomb for non-conductives", look up the term "pump out." Kryonaut suffers considerably from this given its physical properties.

 

[cgeorgescu]

Note one thing: liquid metal makes a huge difference in a powerful laptop. The slimmer the laptop, the more important the cooling is.

• unlike in a desktop (where the cooler is virtually silent), in a laptop you don't want the coolers to spin at max speed because, you know, they sound like a supersonic engine, like the laptop will take off
• at higher CPU temps common in laptops, the CPU spends less time at higher frequencies and scales down because of heat, affecting performance

This doesn't happen easily in a desktop but happens all the time in gaming laptops, especially slim and light ones.

My personal experience going through 2 pastes and one liquid metal application on the same MSI thin and light 17" gaming machine (both CPU and GPU):

• with any paste, CPU and GPU stay in the 85-95C range while gaming, the CPU itself hitting 98C under stress test and benchmarks
• all 3 coolers spin at max RPM, the amount of noise is insane
• metal chassis so, automatically, very hot laptop, could only use it on a table

With liquid metal:

• the CPU temp never exceeds 63C under stress test or benchmarking
• GPU rarely goes over 74C
• no matter how heavy use, even under CPU benchmarking, there's almost no noise from the coolers
• the only cooler noise is under gaming because all 3 coolers are on, but still, no supersonic jet fighter
• due to low temps, my GPU easily runs with cores 200MHz up and VRAM 350MHz up
• due to low CPU temps, my CPU looks like spending way more time at higher temps, my slim laptop gets higher benchmark scores than those large 4kg laptops (with same CPU)

 

[shaolin95]

I feel that I have likely applied thermal compound more than 99% of readers who visit this site from doing cooling reviews as well as the thermal compound reviews. While not possible to get precise to the exact volume each time, my approach is consistent between all applications for both metal and traditional paste compounds. I have built PCs for nearly 25 years, built watercooled PCs for 20 and have been doing the cooling reviews for Tom's Hardware for nearly 5 years, this includes Intel and Threadripper.



I run a 2-hour automated load cycle via batch file which loads Prime95 for 5 minute intervals with 60 second cooldown between processes for each paste application.

This does not fully account for every single paste (addressing the questions around longer cure times) but in doing so also means variables which are not applied to all pastes in the same way, so the attempt here is to make a single process which is applied to all compounds in the same manner. The minimal outliers should not dictate the overall testing procedure.

You did a good job, mate. There will always be a "couch potato big mouth knows it all troll" that will criticize things. Ignore.
Most of us appreciate the time and effort that took to do this.

 

[kOnG lOnG]

Clearly Coollaboratory Liquid Pro is still the original and best liquid metal thermalpaste. Not sure why you're awarding the title of "best" to the lesser option of Thermal Grizzly, but they do like to pay for positive opinion, so I guess I have a clue as to why a product that is 4w worse at transferring heat would still win the title.
Called greed 🤑

 

[BoredErica]

Hope we will see mx-6 tested soon.

 

[JohnDoh337]

Honeywell PTM 7950

Linus Tech Tips finally got around to reviewing it 😆

View: https://youtube.com/watch?v=2BhKx0iQ4K8&t=706s

 

[Avro Arrow]

Well, from the looks of this comparison, for 99.9% of people, it really doesn't matter what you use.

After 30 minutes of a 10-core i9-10850K running all-core Prime95 with an air cooler, no thermal paste allowed the CPU to even reach 65°C. Are some better than others? Absolutely, but except for some power overclockers, who cares?

The way to choose what thermal compound you use would be more about price, safety and convenience.

My thermal compound of choice personally is Arctic Silver Ceramique 2 because it's non-electrically-conductive and you can get great deals on it from Newegg. I bought a 40G syringe for like $25CAD when a ~3G syringe of MX-5 or HT-H2 were $20.

The stuff works great (placing 8th in this test) and for the amount you get at that price, it would be insane to get anything else. I've had it for three years, I've used it several times, I still have at least 1/3 of the syringe left and it never dried out on me. The fact that it's not electrically-conductive means that I don't need to have a heart attack if any of it spills over because it won't short anything out.

One day I came across some Chinese stuff on eBay called "HY-510" and it was so cheap that I bought it just to see if it was any good. I think I paid less than $5CAD for 30g of the stuff. Of course I was thinking, "Yeah, this stuff is probably toothpaste." but I saw an LTT video where Linus bought the cheapest thermal compound that he could find on eBay and the stuff was fine so I thought "What the hell?" and ordered it.

Well it took like 2½ months for it to arrive (which sucked) but it did finally arrive. I had actually forgotten about it by that time because it had been so long. I wondered what method I would use to test it but then the perfect opportunity arose. My stepfather needed to RMA his ASUS X570 TUF motherboard and while it was away, just so that he'd have a working PC, I lent him my backup AM4 motherboard (ASRock X370 Killer SLI) for a few months while his motherboard was on RMA to ASUS.

I figured that here would be a perfect test for this HY-510 because his CPU is a 12-core R9-3900X. He ended up trading me his Wraith Prism cooler (because he HATES RGB) for the Wraith Spire cooler that came with my R5-3600X because it had no RGB on it (Good trade for me!). The Wraith Prism is superior to the Wraith Spire in cooling performance but not to any significant degree. Well, he had zero issues over those three months. I used the HY-510 again when he got his motherboard back from ASUS and that was over a year ago (still no problems). This only solidified my belief that thermal compound is just another chemical commodity in which the differences between brands are essentially meaningless.

 

[letmepicyou]

I believe I still might have a tube of Radio Shack thermal compound...I'd be happy to donate it for testing. 😊

 

[hotaru251]

Your writers need to learn what #1 "best" spot means.

1. ProlimaTech PK-3 Nano Aluminum
Best Premium
Thermal Conductivity: 11.2 W/mk
vs
2. Thermal Grizzly Kryonaut
Runner-up: Best Premium Paste
Thermal Conductivity: 12.5 W/mk



literally only reason it could be "worse" is the cost....yet "best" means you want the best..not the best value.


The best should be the best regardless of price.

Then have runner up being "best value"


this isnt unique to this post. its a recurring trend across all your lists.

Best is the best.
Best is not what you are after if you have a budget.

 

[letmepicyou]

Your writers need to learn what #1 "best" spot means.

1. ProlimaTech PK-3 Nano Aluminum
Best Premium
Thermal Conductivity: 11.2 W/mk
vs
2. Thermal Grizzly Kryonaut
Runner-up: Best Premium Paste
Thermal Conductivity: 12.5 W/mk



literally only reason it could be "worse" is the cost....yet "best" means you want the best..not the best value.


The best should be the best regardless of price.

Then have runner up being "best value"


this isnt unique to this post. its a recurring trend across all your lists.

Best is the best.
Best is not what you are after if you have a budget.

I think they based the #1 on "highest deviation from ambient" rather than "highest thermal conductivity". There are things beyond the thermal conductivity that can lead to lower or higher temperatures, viscosity/how well it spreads under pressure and how thick of a layer it forms thusly, are one. Not only that, but you probably have deviations in how manufacturers rate their products.

 

[jayjr1105]

I use GD900.

works just fine for chips you don't need to OC.

The same company just released GD-2, I just got some in the mail from aliexpress. It's a bit thicker than GD900 but works well so far in my testing.

 

[SyCoREAPER]

No no no. Kryonaut is terrible. It's a benchmarking TP. It dries out quickly and has a thermal limit of 80c before Irreversible drying out occurs

It needs to also be reapplied fairly regularly, months vs years.

Temperatures aren't the only thing that make a paste.

Edit: oh yeah, Kryonaut also has grit in it that cuts/etches/pits the IHS and thermal plate.

I will never use that junk again.

 

[Arbie]

What a huge amount of work! Thanks for the great job and long-term reference.

 

[jaquith]

I'm confused about the data and trying to add or modify testing over different periods of time and different conditions makes the results questionable at best. For example, Arctic MX-5 has been pulled from production whereas Arctic MX-6 is its replacement but is missing?! Any variance within 1°-2° C can be attributable to nearly an infinite list of variables.

 

[Avro Arrow]

I believe I still might have a tube of Radio Shack thermal compound...I'd be happy to donate it for testing. 😊

Which brand is it? Archer, Realistic or Optimus? 😀

 

[mopeygoth]

Strange how you can say that a paste with aluminium in it does not have any drawbacks, it will completely miscolor a copper heatsink over time. It is irreversible.

Also i thi k your Arctic silver has gone bad 😛

Howcome you are not testing mx6? The most hyped paste of 2022

 

[chucklepard]

We tested 90 different thermal pastes to help you find the best one to help cool your CPU.

Best Thermal Paste for CPUs 2021: 90 Pastes Tested and Ranked : Read more

Several of the reportedly best performers (#5, #9, #10) are showing relative effectiveness of 1, where one means horrible and 5 means outstanding. Is this simply a reoccurring typo in the review, or is there some subtlety I am missing here?

 

[mac_angel]

I was a little surprised not to see Thermalright TFX Thermal Compound. I've been seeing people mention it in a lot of forums.

 

[LoneWolf_53]

I must say I'm a tad baffled by some of the test results as they don't strike me as reliable.

For example TG Kryonaut is listed as having a thermal conductivity of 12.5 while Arctic MX5 is less than half of that at 6.0, yet the effectiveness results are both listed as 4.

Does it not stand to reason that double the thermal conductivity should have a significant better outcome than the lower rated product?

Something is amiss.

 

[Digital~Dreams]

My 12900ks missing from Toms processor charts as is the Thermal Grizzly Kryonaut 'Extreme' paste I use !?!?

 

[vinay2070]

Hope we will see mx-6 tested soon.

I second this.

 

[niceguy1]

We tested 90 different thermal pastes to help you find the best one to help cool your CPU.

Best Thermal Paste for CPUs 2021: 90 Pastes Tested and Ranked : Re
What about GD900 paste? Seems to be better than noctua nt-h1

We tested 90 different thermal pastes to help you find the best one to help cool your CPU.

Best Thermal Paste for CPUs 2021: 90 Pastes Tested and Ranked : Read more

 

[Deleted member 431422]

Too bad you didn't get to test 7950X with the thermal paste. I remember reading the temperature under load is around 90°C. It'd be interesting to see if what the difference with top cooling compounds would be.

 

[Avro Arrow]

"We tested each thermal paste with a low-tension air cooler mount, high-tension air cooler mount, and a high-tension AIO liquid cooler mount. Each application was given a 1-hour burn-in using Prime95 with ten load and cooling cycles over the course of the hour; six minutes each with a ten-second cool-down between. Each testing load session was then executed for a one-hour load period, again using Prime95."

Well, if the hottest that the "worst" thermal paste in the test made the CPU in a test like that is only 65°C, then the correct answer is that it doesn't matter what you use as long as you use thermal paste. Let's face it, 65°C is very comfortable for a CPU and nobody runs Prime95 all day. Unless you're some extreme overclocker (in which case you use LN2 or LHe anyway) the only thing that should make a difference to you is the cost of the stuff.

I use this stuff I got on eBay called "HY-510" and it works just fine. I've used it on my dad's R9-3600X, my own R5-3600X, R7-5700X and R7-5800X3D, all without issue. I've read that it's comparable to Deepcool Z9 which, at most, had the CPU running only 6°C warmer than Thermal Grizzly Conductonaut.

I honestly think that HY-510 is the best thermal compound for most people. We see here that it's within 2°C of Arctic MX-4:

and it only costs 8¢ per gram.