If geeks love it, we’re on it

Thermal Paste Mini-Roundup!

Thermal Paste Mini-Roundup!

Performance Results and Conclusions

Please see our ‘Application and Testing Methodology’ page for details on how we tested. A picture is often worth one thousand words, so without further ado:

Although there is not much of a difference between these pastes, there is indeed a measurable difference. Whether or not a one degree delta is enough to sway a buyer from one product to another is difficult to say.

What was most surprising was how poorly that old no-name silver compound I had compared to the others. I expected maybe a degree or two, but a full 5.5°C delta between it and the TX-2 is pretty incredible. So there you have it! Some surprises and some of what I expected. There have definitely been some improvements in the composition of these thermal interface materials over the last several years. The fact that these four pastes were able to best the already potent Arctic Silver Ceramique by a full degree is impressive.


One particular paste really stands out from the crowd—Tuniq’s TX-2. Not only is it the best performing paste of the lot, it is the cheapest! What more could you ask for? For a measly $3.99 per tube, this stuff is a steal. If I had to buy some paste for a personal machine of mine—it would definitely be TX-2.

OCZ’s Freeze is also a very potent thermal interface material. It managed to best our reference ceramic-based paste by a 1.5°C margin. It can be found at just about any PC retailer and is priced reasonably as well.

Coolink’s Chillaramic compound comes in a massive 10g tube, making it the most plentiful of the lot. This is definitely a great choice for those who do a lot of CPU installations. Its performance is right up there with the best and has a full degree on our reference ceramic paste.

Noctua’s NT-H1 is another great premium paste. Buyers of Noctua’s recently released heatsinks can rest assured that the included NT-H1 is a top-notch product. It also bested our reference ceramic compound by a full degree. Like the other pastes we tested today, it is non-conductive, and does not require burn-in for best performance.

I was very happy with all four of the products I tested today, and am pleased to provide awards all around. Tuniq’s TX-2 is by far a clear best choice product and I’m happy to award it with Icrontic’s rare “Golden Fedora” award. The OCZ, Coolink and Noctua pastes are also fantastic products and definitely deserve the “Icrontic Outstanding Product Award”. A big thanks goes out to Tuniq, OCZ, Coolink and Noctua for providing us with these samples.

Discuss this review in our forums.


« Previous

Comments

  1. Leonardo
    Leonardo Thanks, Mike. Expect to get complaints that you did not apply each paste in accordance to a particular guide posted hither or yon. I think your methodology was sound, testing each the same way and taking pains to be consistent. That's what it's all about - comparison.

    BTW, I'm quite sure why the article states that TX-2 has been out for only six months. I've been using it for over a year.
  2. Zuntar
    Zuntar Nice review Mike, solid and to the point without all the fluff that some sites go for!

    Thanks for your hard and consistent work!!
  3. lemonlime
    lemonlime Thanks for the comments guys.

    I was a bit concerned about that too, Leo. But like you mentioned, consistency is most important when comparing paste products. With half degree deltas, I had to rely on an application method that I could reproduce over and over again. The problem with the bead and line method is that pastes with varying thickness may spread differently under mounting pressure. Spreading it as described allowed me to see exactly how thick the applied paste was.

    You are right about the TX-2 being around for over a year. Our PR rep contacted us about six months ago regarding TX-2 and I incorrectly assumed it was freshly released at that time. I corrected that statement in the review. Thanks for catching that! :)
  4. jokerz4fun
    jokerz4fun Great job Mike, I always thought they were all the same.

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!