Thermal pastes from the legendary Arctic MX series are getting a new addition with an updated formulation: Arctic MX-7. Compared to its predecessor (MX-6), it is characterized by higher viscosity while simultaneously delivering higher cooling performance—at least according to internal tests published by Arctic. The new Arctic MX-7 thermal compound is available in three size variants, in tubes containing 2, 4, and 8 grams of material.
Based on Arctic’s testing, the new MX-7 thermal compound should deliver higher cooling performance with the same cooler than the older MX-6. The differences are not dramatic, but some improvement is expected—the temperature of the test processor, an Intel Core Ultra 9 285K running at approximately 284 W, dropped by nearly two degrees Celsius (likely an average across all P-cores).
These results point to higher thermal conductivity, which Arctic does not explicitly list in the specifications. Indirectly (based on the temperature differences), however, we learn that it should be higher. By how much is difficult to say. In any case, the results of temperature tests across generations of Arctic MX thermal compounds are shown in the chart below.
Arctic lists the viscosity in the range of 35,000–38,000 Poise. The consistency of the paste should therefore be extremely fluid. This is typically beneficial in that the material can reach even the smallest gaps that thicker pastes cannot, ensuring contact over a larger surface area—which naturally results in faster heat transfer. This is the theory, and whether it holds true in practice for Arctic MX-7 is something we can only trust for now. The stated density, expressed in grams per cubic centimeter, is 2.9 g/cm³.
The consistency is said to be so fluid that the use of a spatula is not recommended when applying MX-7 thermal paste. In this respect, MX-7 is closer to MX-6 than to MX-4, which is also illustrated by the axis below.
It may be suitable to apply the thermal paste in a cross pattern, ensuring that the corners of the processor’s IHS do not remain “dry” and without contact, even when using a smaller amount. This cross-pattern technique proved effective in our own tests in the past. In this context, Arctic also mentions the use of the thermal compound directly on silicon dies, which is relevant for graphics cards or laptops and other mobile devices with higher heat-transfer requirements.
There are four different packages across three tube sizes: 2, 4, and 8 grams. The 4-gram version is also available as a bundle that includes cleaning wipes. Retail prices (MSRP) are as follows: 14.50 EUR (2 g), 16 EUR (4 g), 17 EUR (4 g with six MX wipes), and 21 EUR (8 g).
Based on previous experience with Arctic products, lower prices—often roughly half—can be expected in stores. Some discounts will likely also be available when purchasing directly through Arctic’s e-shop.
English translation and edit by Jozef Dudáš
Igor’s Lab test showed it to be very good paste with near top performance, while there are improvements for durability that may indicate better prognosis for pump out effect, he clearly stated this something that needs to be tested in real tests and he will carry them out in the future. I have several tubes from Arctic, MX5 sadly was a failure and was discontinued quickly, MX6 is quite a decent paste but still not very stable for direct die usage (I remember testing it on delided 12900K and within few days performance deuterated rapidly). I hope MX7 will be more stable. I know every paste that doesn’t very high viscosity (like Thermalright TF-X which is like concrete and is PIA to apply) will have pump out effect the question is the severity, I’m ok repasting every year to year and a half but not every few months.
Thank you for sharing your user experience. If you decide to go with MX-7 at some point, we would definitely appreciate it if you shared how you perceive this thermal paste—especially after longer-term use. 🙂
The original plan was to include our own illustrations in the article. Arctic even sent us one tube for that purpose, but it took longer to reach the test lab than expected… and I’m adding one photo only here, in the comments. The paste doesn’t actually look that thin, does it? Of course, temperature also plays a role—here it’s about 21 °C. 🙂