The questions about the right thermal paste can quickly lead to controversial discussions in pc hardware forums. One person swears by the expensive premium paste, while another has been using a cheaper option for 20 years and is just as convinced of their choice.
As a regular Arctic MX-4 user, I have been deeply interested in this question myself. So, I set out to find the best thermal paste for my gaming PC and provide you with all the necessary information to potentially optimize the thermal coupling between your CPU and its cooler.
Basics of Testing Thermal Pastes
In the past, thermal paste manufacturers often advertised the "W/mK" value (watts per meter and Kelvin). However, since this is purely a laboratory value and not standardized, it says little to nothing about the actual performance of a paste in a PC. As a result, most manufacturers have stopped listing it altogether. Incidentally, this is something that CPU and cooler manufacturers could take note of for the TDP value, but that’s another topic.
So, how do we compare the performance of different pastes? The best way is by measuring the CPU temperature in the context of a real use case! Therefore, I turned my gaming PC into a test bench to obtain comparable and reproducible results. More about the test system and methodology later.
"Break-In / Burn-In" and "Pump-Out"
The myth of "break-in" (less frequently called "burn-in") has persisted stubbornly for many years. This refers to the idea that thermal paste needs several dozen or even hundreds of heating and cooling cycles to reach its full potential. Arctic Silver 5 paste from Arctic Silver Inc. (not to be confused with Arctic GmbH, responsible for MX-4 and MX-6) is one such example. The YouTubers "Tech Illiterate" and "ponchato" have examined the "break-in" period in their videos (1, 2) and both concluded that there is no performance improvement after a "break-in" period. Instead, the thermal conductivity of the pastes remained the same. Noctua even prints a note on the packaging of their thermal pastes stating "No break-in or burn-in required."
So, what is "pump-out" and how does it relate? "Pump-out" refers to the squeezing out of thermal paste from the center of the processor to the sides. This occurs when the CPU and cooler heat up and expand under load, then cool down and contract when idle. These changes are microscopic but sufficient to displace cheaper pastes from the center to the edges, compromising thermal transfer. Arctic explains this behavior on their YouTube channel to promote their MX-6, which supposedly does not exhibit this behavior.
In my tests, I observed the pump-out effect to a minor degree in almost all pastes. The first measurement was always slightly cooler than subsequent measurements. However, following measurements remained very consistent.
The Candidates - From Low-Budget to Extreme Overclocking
First of all, a big thank you to Arctic, Noctua, and Thermal Grizzly for providing their thermal pastes free of charge for this review! None of the manufacturers imposed any conditions on the review. It's great that you support small creators like me with samples! I procured the paste from Thermalright myself. This is mentioned here for the sake of transparency.
Thermalright TF7: Affordable Yet Capable?
At just over 3 euros for 2 grams, the TF7 from Thermalright is a very affordable thermal paste. However, it is often said to not perform as well as the usual suspects (like Arctic MX-4). Perhaps I can clear up this prejudice here.
The TF7 is rather difficult to spread, even with the help of the supplied spatula. It is quite viscous and tends to tear. Its most compelling argument is clearly its very low price and its inclusion as a bundle with common Thermalright coolers. Is this paste sufficient to establish good contact between cpu and cooler?
Thermal Grizzly Aeronaut: Premium Ease of Use
Priced at €7.49 for a 1-gram tube on Amazon, the Thermal Grizzly Aeronaut is positioned in the higher-end segment. Compared to the previously mentioned Thermalright, it is extremely easy to work with. Especially for beginners who have little experience with applying thermal paste, the Aeronaut's more fluid consistency is a big advantage.
I also appreciate that it comes in a resealable zip bag, which should protect the paste from drying out. The 3.9-gram tube provided for this test even includes two small screw-on applicators. This makes it very easy to apply and spread the paste effectively.
Arctic MX-4: The Classic Thermal Paste
The absolute classic among thermal pastes, the MX-4, is priced at €4.50 for a 4-gram tube. While initially more expensive than the Thermalright option, it becomes more cost-effective when calculated per gram. The MX-4 spreads well, although it falls short of the Thermal Grizzly Aeronaut in this regard. It also tends to string, which makes handling a bit more challenging.
I've been using Arctic MX-4 for over 10 years and have always been very satisfied with it. This is why I'm particularly curious to see how it fares against the high-end pastes in this roundup. As a little extra, I also included a tube of MX-4 that I bought back in 2019 in this comparison.
Arctic MX-6: The New High-Performance Contender
Staying with Arctic, let's take a look at the MX-6, the manufacturer's high-performance thermal paste that was released in 2022. Priced at €4.09 for a 2-gram tube, it isn't too far off from the MX-4 and TF7 in terms of cost but claims to offer 20% better performance than its sibling.
Due to its high viscosity, spreading the MX-6 is somewhat more challenging. The paste tends to tear easily, making it less user-friendly for beginners compared to the Thermal Grizzly Aeronaut. I'm very curious to see how it performs in temperature comparisons against the MX-4.
Noctua NT-H1: The Veteran Thermal Paste
Unlike the MX-6, the Noctua NT-H1 is a veteran among thermal pastes. It has been around since 2007 and is frequently recommended in forums. "Old but gold" is an apt description, as its almost creamy consistency makes it exceptionally easy to apply to the CPU. Like the Thermal Grizzly Aeronaut, the NT-H1 is an absolute beginner-friendly paste in this regard.
However, with a starting price of €9.90 for a 3.5-gram tube, it is on the expensive side. But if you consider the small sealing frame that you can optionally choose for the same package price, the NT-H1 becomes an even more attractive package.
Noctua NT-H2: Aiming for the Top
When Noctua introduced the NT-H2 in 2019 as the successor to the NT-H1, their goal was to compete at the highest level once again. Positioned in the premium segment, the NT-H2 offers a more moderate price compared to Thermal Grizzly's premium offerings. At €13.90 for 3.5 grams, it’s quite reasonable.
The application of NT-H2 is more suited for advanced users. Despite being somewhat challenging to spread, it tends to string more than any other paste in this test, leaving a small splatter on my motherboard right away. But don’t worry, like all the other pastes in this comparison, the NT-H2 is non-conductive. Conveniently, Noctua sells it bundled with three or ten cleaning wipes. I tried one and found them to be pleasantly large, sufficiently moist, and generally easy to handle.
Thermal Grizzly Kryonaut: The Pinnacle of Performance
Thermal Grizzly Kryonaut is synonymous with performance. Developed for extreme overclocking, it even boasts a "Liquid Helium -250°C approved" logo on the packaging. But does it bring this level of performance to a typical ITX gaming PC?
Kryonaut is by far the most expensive paste per gram in this test. However, the 1-gram pack is priced at €7.90, making it cheaper than the smallest 3.5-gram pack of Noctua NT-H2. How this compares in terms of value will be discussed later. Initially, I found that this paste is also very viscous. Like the NT-H2, it tends to string extensively. Therefore, I would say this paste is better suited for advanced users due to its challenging application. Using the provided applicator, which screws onto the front of the tube, the spreading process becomes more manageable, though.
Test System and Methodology
As mentioned earlier, my own gaming PC serves as the test system. However, I learned from my previous test of SFF air coolers (see link at the bottom) and measured the temperatures using an open test bench this time. This way, I don't have to repeatedly extract the cooler from my small FormD T1 case.
Test System Specifications
CPU | Ryzen 7800X3D * @ 76,7W | CO -20 |
CPU-Cooler | AXP90-X47 Full * @ 2600 rpm |
Fan (directed airflow to CPU cooler) | 1 x Phanteks T30 * @ 980 rpm |
Graphics Card | |
Motherboard | |
Power Supply Unit |
Two thermometers, which measure the ambient temperature, are placed directly in front of the Phanteks T30 fan, which directs air towards the AXP90-X47 Full cooler.
Each thermal paste was tested in at least three runs of 10 minutes each using Cinebench R23. The respective temperatures were recorded, and the averages were calculated. The ambient temperature was then subtracted from these averages, and a temperature value of 23.5°C was added back to ensure comparable results.
Of course, with every change of thermal paste, I checked the contact pattern of the cooler on the CPU before I cleaned everything and then used a small spatula for the next application, like the one included with the Thermalright TF7. This was done to create even coverage on the CPU heat spreader. In terms of quantity, I did not skimp and preferred to apply a tad more, so that a little bit of paste was always pushed out over the edge of the CPU heat spreader.
The Results - Which thermal paste is the best for your gaming PC?
Looking at the hard numbers, it’s clear that there are no gigantic differences like when comparing paste to liquid metal for example! The three best pastes in the test, the Arctic MX-6, the Thermal Grizzly Kryonaut, and the Noctua NT-H2, are about 1.4 to 1.7°C ahead of their less strong competitors.
A clear trend can be seen. The three front-runners are so close together in the average temperatures from three runs, with just 0.3°C difference, that we can probably speak of error tolerance here.
I personally noticed that the NT-H2 always had a short temperature peak in all runs, but otherwise performed at the same level as Kryonaut and MX-6 99% of the time.
I find the result that the Noctua NT-H1 was able to achieve interesting. It is technically in the middle of the field and has already stood out due to its easy application. The three "High-Performance" candidates mentioned before are a bit more demanding on the user’s application skills.
At the end of the test field, the Arctic MX-4, Thermal Grizzly Aeronaut, and Thermalright TF7 are set. If the three did not differ noticeably in their spreadability, I would have suspected that it must be the same paste, as the temperatures are almost identical. We will come to the classification of all results at the end in the conclusion.
5-Year-Old Thermal Paste - Can you really still use it?
First, of course, I want to show you my Arctic MX-4 from 2019 again in comparison with the fresh tube. Unlike Noctua, which prints a “recommended storage time before use” of three years on the packaging, Arctic does not provide any information on this. So what can the 2019 MX-4 do in 2024?
I absolutely did not expect that. The measured values are almost identical. On paper, even slightly better, which strengthens my assumption that a small error tolerance should be assumed for all measurements. The spreadability of the 2019 paste is also analogous to the new tube that Arctic sent me for the test. After just over 5(!) years of irregular use, the MX-4 still delivers its full performance. Impressive.
So don’t necessarily let yourself be fooled by the disclaimers of the manufacturers or panic-mongering in any forums. If you still have an older tube of thermal paste, don’t necessarily throw it away. Take a close look at the paste and if it hasn’t become oily, dried out, or otherwise decomposed into its components, just use it and see if the temperatures still fit. My sample here shows that an old paste can still offer exactly the same performance as a brand new one.
Which thermal paste is the cheapest per gram and how many grams should you buy?
Now that we know which pastes have delivered the best results and how expensive they are compared to the competition, can’t we now create a price-performance chart? Well, it’s not that simple, but let me explain.
First of all, it should be noted that there are no uniform packaging sizes for thermal pastes. Thermal Grizzly’s smallest tubes contain 1 gram, Arctic and Thermalright go with 2 grams, and Noctua’s pastes even start at 3.5 grams. This makes prices difficult to compare. But wait, what about the price per gram? This also varies depending on the size of the tube, so it is not a uniform value either.
Let’s try a practical approach: For a job on the CPU, i.e., one application, I needed about 0.2 grams of thermal paste in my test. Sometimes more, with the very viscous pastes like the NT-H2 or the Kryonaut, sometimes less with the more liquid pastes like the Aeronaut or the NT-H1. So if you build a PC once and then want to renew the thermal paste after about 3 years, even the small 1-gram tubes from Thermal Grizzly should be completely sufficient.
Are you a PC enthusiast like me and regularly tinker with your computer? Then I recommend you to have at least 3.5 grams of paste in the drawer. There’s nothing worse than wanting to mount your CPU cooler and running out of thermal paste (this has actually happened to me before).
An then, if you frequently build PCs as a side job or for friends and relatives, you might also consider the 20 or 45 gram tube of Arctic MX-4. Thermal Grizzly sells the Kryonaut in a 37 gram tube, which is likely to be a popular choice for hobby and extreme overclockers.
I have tried to clearly present the pricing policy of the various manufacturers in a table for you. The prices are as of May 2024 and were taken from the price comparison website Geizhals.de. The highest prices are marked in red, the lowest in green. The upper value always shows the entry price, the lower value relates to the price per gram.
The Conclusion - which thermal paste would I recommend to you?
If you have read the article carefully and compared it with your personal requirements, you may have already found the answer for yourself. For my part, I am enormously impressed by the price-performance ratio of the Arctic MX-6. Not only can it keep up with the Kryonaut, which is twice as expensive at the entry price, and the NT-H2, which is three times as expensive, in my test, but it can even be spread a little easier than the other two competitors. For my air-cooled Small Form Factor (SFF) system, it is therefore a great paste at a great price!
It is very possible that a paste like the Kryonaut performs a little better on a CPU with higher heat output, such as a 14900K, which also has a custom water cooling system, but in my specific test case, both pastes delivered almost identical performance between the CPU and cooler. The MX-6, as mentioned, has the clear price advantage.
The Noctua NT-H2 also doesn’t need to hide. Only its somewhat high purchase price for the already rather large 3.5 gram tube stands in its way for occasional builders from my point of view. Thermal Grizzly does this better, I think, and positions both the Aeronaut and the Kryonaut with the 1 gram packages in the range of 5-10€ next to the other pastes, which all offer more content, but as we have already determined above, you may not need at all if you only want to build a single PC.
Thermalright only offers the TF7 in one size. It has performed identically to the Arctic MX-4 and the Thermal Grizzly Aeronaut, but two things stand in its way: On the one hand, it is rather tough in consistency and thus quite difficult to process, so not necessarily something for beginners. On the other hand, the tube of the Arctic MX-4, which is twice as large, is just a few cents cheaper. Nevertheless, the many cries of doom that one reads about this paste are unjustified! If you use a Thermalright cooler, which comes with this paste, you don’t need to reach for an alternative, but can use it with confidence and expect good results.
What conclusions can we now draw from the above information? I will try to highlight the perfect thermal paste for the corresponding user type from my point of view.
One Time PC Builder - The Beginner
You build a new PC every few years, or perform an upgrade, and are actually quite glad if everything is as easy to handle as possible? The perfect pastes for you are the Arctic MX-4 because of its legendary low price and longevity, the Thermal Grizzly Aeronaut, due to its extremely easy application with the included spatula, and the Noctua NT-H1 (including the AM5 gasket), which can even add a small performance plus compared to the other two but is still super easy to spread on the CPU.
On the other hand, an extra 10-20€ at the end of a 2000€ PC isn’t that significant, right? So if you’re willing to pay this price for 1-2 degrees lower temperatures, then you can of course also look around in the enthusiast segment.
PC Enthusiasts and Tinkerers
Are you an advanced user who likes to tinker with their PC and wants to get the most out of your machine? Then you’re on the right track with the Arctic MX-6, Noctua NT-H2 or Themal Grizzly Kryonaut! Personally, I would probably go for the 8 gram tube of Arctic MX-6 and wouldn’t have to buy any other paste for many months, probably even years.
However, I must also admit that I have already used Thermal Grizzly, Thermalright, and Noctua pastes before this test and therefore did not follow my own advice. I just enjoy trying out different products. Maybe you also have two or more different thermal pastes in your drawer?
Regular PC Builders, PC Shops or PC Building Services
Do you have your own PC shop, or are you perhaps in charge of hardware in a small company’s IT department? Are your friends and relatives knocking down your door for you to maintain and update their computers? Then the Arctic MX-4 in the 45 gram pack is probably the right paste for you. At only €0.31 per gram, it is unbeatably cheap and still a very good performer for the average user.
Which thermal paste do you use?
In the end, I’m interested in which paste you swear by? What gives the best performance in your PC and have you ever tried other manufacturers products?
Please feel free to leave a comment with your experiences under this post!
Thank you for reading!
If you liked this post and want to read more comparisons like this, feel free to check out the linked posts below or follow the button to my Ko-Fi to become a supporter.
*The links marked with * (asterisk) are so-called affiliate links. If you make a purchase through one of these links, I receive a small commission. The price for you does not change as a result. Thank you! :)
Comentários