Thermal Hero Thermal Paste

Details: Category: Hardware; By Jason Gress; 05.Apr; Hits: 1523

Game Info:

Thermal Hero Thermal Paste

Specifications:

Thermal Hero Neo
Thermal Conductivity: <=12 W/mk
Density: 2.8 g/cm^3
Thermal impedance: 0.009 C-in^2/W
Operating Temp: -50-220C
Viscosity: 215 Pas

Thermal Hero Ultra
Thermal Conductivity: <=13 W/mk
Density: 3.45 g/cm^3
Thermal impedance: 0.003 C-in^2/W
Operating Temp: -50-300C
Viscosity: 240 Pas

Thermal Hero Quantum
Thermal Conductivity: <=17 W/mk
Density: 3.5 g/cm^3
Thermal impedance: 0.0026 C-in^2/W
Operating Temp: -75-300C
Viscosity: 230 Pas

Thank you Thermal Hero for sending us 2g each of these thermal pastes to review!

In the process of attempting this review, I have learned just how difficult reviewing thermal paste actually is. I recorded some runs of Prime95 and Cinebench R23 with HWiNFO64, and thought I was good to go, until I realized that my prime95 numbers are all but invalid, because I let some run longer than others (probably on accident), which would normally be fine, but the hottest spots showed up later than I let some runs go. So I have to rely on Cinebench R23 temperature readings in order to keep things consistent.

Speaking of which, when setting up for the final test, my CPU got stuck to my cooler's cold plate, and pulled out of the socket. This unfortunately led to my highly tuned overclock being lost once I plugged it back in. I was able to restore most of it, but I do not believe it's exactly the same as before. So my settings were probably slightly off on my final test, and I can't go back because I received only 2 grams of thermal paste for each sample, and that 2g was all but used up for a single application, so I cannot redo my tests. My CPU got stuck to the cold plate with both the Ultra and Quantum paste; this did not happen with the Neo or the existing Arctic pastes.

All of that to say, I did my best, but these tests are not perfect. I still hope you are able to gain some insight from my findings, though.

Before getting into my results, I should probably get into more detail on my test setup. My day-to-day setup is admittedly a torture test of sorts, where I run an AMD Ryzen 5950X CPU, with PBO (Precision Boost Overdrive) enabled, along with tuned RAM and a mild curve optimizer setup on an Asus Crosshair VIII Wi-Fi. (There may have been other tweaks, but again, those settings are lost.) My heat sink is a five year old Corsair H100x, 240mm all-in-one (AIO) water cooler. I run this inside of a Fractal Design S2 Vision RGB, which is a wonderful-looking case with a decent design, but flawed thermals. I mitigated this somewhat by using the mesh top instead of the glass top, but the front still limits air intake. I run 32 GB of Corsair Dominator DDR4 running at 3666MHz at CAS 16 (I lost the silicon lottery, despite purchasing Samsung B-Die RAM). My case fits inside of a cubby in my desk, where I added a 140mm and an 80mm fan, pulling that trapped hot air out of the cubby. The front of that cubby is open, but this setup is far from ideal for good temperatures. Thankfully, because of all of those fans and large heat sinks, I'm still able to maintain good temperatures despite all of this working against me.

I ran five tests, first before any changes, then with four different thermal pastes. The first is my standby recommendation, Arctic MX-6, and the three review pastes, Thermal Hero Neo, Thermal Hero Ultra, and Thermal Hero Quantum.

Highlights:

Strong Points: Easy to apply and clean up; Quantum has excellent performance, and Ultra is a decent value
Weak Points: Neo has terrible thermal performance; they have strong suction with the CPU, enough that my CPU was pulled out of the socket

When applying the paste, I drew a 'plus' with lines in the middle, and dotted the four corners to make sure that area was also covered when pressure is applied. I know I wasted some of the paste, and if I was careful I might have some left over for another application, but I can say with confidence that the entire heat spreader on the CPU was covered cleanly, as I removed the AIO cold plate after I was done and saw that there was plenty of thermal paste to go around. I probably used too much, but when it comes to thermal paste, too much is definitely better than not enough.

Below, I list my Cinebench R23 scores, maximum power draw, and maximum temperature as measured by HWiNFO64.

Baseline (probably Arctic MX-4):
Cinebench R23 score: 28072
Max CPU Package Power: 221W
Max Core Temperature Average: 77C

Arctic MX-6:
Cinebench R23 score: 28531
Max CPU Package Power: 222W
Max Core Temperature Average: 77C

Thermal Hero Neo:
Cinebench R23 score: 27915
Max CPU Package Power: 212W
Max Core Temperature Average: 82C

Thermal Hero Ultra:
Cinebench R23 score: 28266
Max CPU Package Power: 216W
Max Core Temperature Average: 79C

Thermal Hero Quantum: *results after CPU settings reset
Cinebench R23 score: 28514
Max CPU Package Power: 226W
Max Core Temperature Average: 76C

Now it has to be said that I didn't get my system perfectly stable until a while after the last measurement, but these results are still useful because the power output and temperatures are still valid measurements.

A few things to note, in addition to the straight numbers above. It has to be said that the Thermal Hero Neo paste is very thin, and does not work well, at all. It was the only paste to hit TDP (Thermal Design Power) maximum while testing with Prime95. It also has the worst thermal performance in this test. I do not recommend this paste.

The Thermal Hero Ultra paste appears to be their midrange paste, and it's decent. It doesn't perform quite as well as Arctic's MX-6 in my limited testing, but it's not terrible. It's currently on sale on Amazon for a very good price, so if you need a lot of paste for a good price, I can safely recommend it. It's an easy to work with paste, and is fairly thin and easy to spread.

The Thermal Hero Quantum paste seems to be their top of the line for non-conductive pastes (They sell a liquid metal, which my heat sink is not compatible with.) The performance appears to be a step above everything else I tested, so I can recommend it based purely on performance for sure. It's a dark gray paste, and while thicker, it's still easy to spread and work with. Whether it's a good value depends entirely on price; at current Amazon prices, I'm not sure it's worth the steep increase in cost. But if there is a good sale, or you are willing to buy a large amount at once (a 25g tube is a fair price, if you need that much) then it perhaps could be worth it.

It has to be said that the Arctic line of thermal paste has been my go-to paste for many years, as it's always worked excellently. And for the price, it's still a solid choice. At a price of around $1.50 per gram, it performs well and is a great value. Thermal Hero's Neo is only a better value with their 50g pack, which brings the price down to just $1.24 per gram, which is great if the performance drop is worth it to you. Thermal Hero's Ultra is currently on sale for an insanely low $7 for 10 grams, and totally worth it at that price. While the performance is slightly below the Arctic in my testing, it still offers good value.

For the best performance in a non-conductive thermal paste, the Thermal Hero Quantum is pretty great. But the price is quite steep. The best value currently is 25 grams for around $2.72 per gram, and the other tube options are way overpriced in my opinion. I'm guessing this paste is meant to compete with something like Thermal Grizzly Kryonaut, which is priced in the same range as this Quantum, and is also billed as a high-performance paste. I have used that before; it's a huge mess to work with and I would easily pick the Quantum paste over that.

I learned a lot from my first review of thermal paste. I didn't even share all of my trials in this process; I had to use a razor blade to straighten bent pins on my CPU from one of the unexpected CPU removals! Despite all of that, hopefully this comparison was worthwhile. If you need thermal paste, consider your needs. If you intend to overclock or work with higher-end CPUs that put out 200+ watts, then I recommend buying higher-end thermal paste. When you want the best, you go for the best. Thermal Hero Quantum is a solid choice. If you are looking for a great value for most common uses of thermal paste, the Arctic products or the Thermal Hero Ultra is a good choice. I do not find the loss in performance measured with Thermal Hero Neo worth considering, if it were for my electronics.