I've been using Arctic Liquid Freezer III (not Pro) for a while now. When the Pro product line was released, I immediately replaced the original P12s with Pros, and I was very satisfied with the new cooling performance and the noise level. I believe the cooler now considered as a Pro, as the major difference between the regular and Pro cooler is only the fans.
Then a question crossed my mind: what other similar simple modifications could be made to improve or optimize the cooling performance, and how much they actually mean in numbers. And this comparison could be a fun project. Of course, this test covers only one given hardware, one type of stress test and many other options are left out… but it is still instructive, at least for me.
I had prior expectations of what would really matter and what would not. There were surprises in the end.
Since the CPU (13700K) has been undervolted for quite some time, I also included this factor in the comparison. I found my test results and measurements from the initial, basic settings, from the time when undervolting wasn’t done. As I expected, the most significant optimization potential is here, it is not worth skipping this step. So the undervolted CPU with stock AIO became the real reference point to which I compare all other versions.
Since I wanted to test results that can be used in real life when noise level is also important, most versions use exactly the same fan curve tuned for quiet operation. Just for the sake of the test, I also looked at maximum speed too.
Before the test, I thought that a significant improvement could be achieved with stronger fans or more fans, but I was wrong. Neither the max speed nor the pull-push arrangement made a measurable difference in the temperature. Neither the fan replacement. At the moment I’m using the system with the T30s, but cannot see or hear any major difference to the Pro setup.
On the one hand, the Pro fants are very good in absolute terms, not only in price-value, on the other hand, the Arctic AIO seems to have been designed quite well and balanced. With the Pro fans it has no strong or weak points. Of course, it would be worth comparing the 240 with the 360 or 420. Maybe next time. The 360 version will just fit into my regular Fractal North case.
The last and most complicated modification was replacing the OEM contact frame to Thermalright Contact frame. This can't be done just like that but we require a mounting frame. There is no official mounting kit for LGA 1700, but the Arctic LGA 1200 kit actually works very well, but you must have a contact frame to avoid insufficient pressure on the IHS.
Since I had already removed the coldplate-pump block, I thought I would try a thermal paste that was considered good. MX7 paste wasn’t release yet that day when I made the test, so I just picked Kryonaut. Using liquid metal would have been the best but it seemed too difficult to me. I was thinking on testing PTM 7950 thermal pad but I read that it works well if we have higher temperatures, and my CPU operates usually below 50C. So I just picked something that an average PC owner would pick😊