Mounting the cooler was quite a task. In order to do so, you have to remove the motherboard from the case. Above you can see the plate which goes on the underside of the motherboard. You then need to apply thermal paste to the CPU and screw in four screws from the underside of the motherboard, in to the cooler itself. This is particularly difficult, purely because of the size of the cooler.
Once affixed, the cooler isn't going anywhere. As it weighs close to 1kg, MACS has actually supplied cables that can be fitted in your case to take some of the strain. As we were using an open aired test-bed, this wasn't necessary. But it's a nice idea, as too much stress on a motherboard can cause micro-fractures in the traces on the PCB.
Another issue I came across when fitting the cooler, was the connector for the TEC. It can fit in to the control unit either way around. The first way I plugged it in (which looked right), was incorrect and I noticed that the heatsink was getting cooler instead of warmer. I then quickly turned it around, but someone less experienced could have very quickly cooked their CPU to a crisp.
The manual wasn't as good as it could have been, with no troubleshooting section, or FAQ. The web site was no extra help either. A forum would have been nice, for at very minimum discussing issues with other people who owned the product.
We tested the cooler with a Core 2 Extreme QX6700, which is a quad core processor. None of the Core 2 range run particularly warm, but this is without a doubt the warmest, purely due to the number of cores.
Using the MACS cooler, we overclocked the CPU to 3.73GHz (from 2.66GHz) using 1.55V. It wasn't quite stable in all tests, suggesting that one of the cores wasn't happy. However, it was still a good way of measuring cooling ability.
At this speed, the CPU was 64c at 100% load on all four cores. For comparison, the incredibly loud cooler that Intel supplied also ran at 64c. However, from an efficiency point of view, the MACS cooler was superior, as the surrounding (and therefore incoming) air was 54c, while the standard air cooler had air at 33c.
This does however, illustrate a very important point. The TEC inefficiency means there is a lot more heat being generated on the other side of the CPU. Ours was an open test bed and it still couldn't cope with the extra heat. This means the VRMs and graphics card were significantly warmer than normal. In a case, temperatures would escalate very quickly indeed. If you are intending on running this cooler – you better have decent enough case cooling to cope with the extra heat generated.
If (and that's a big if) you can remove all the extra heat being generated, the high efficiency of this cooler means that you could lower CPU temperatures considerably. Traditionally, and for good reason, water cooling is still the best solution for this, as the heat is taken outside of the case.
MACS have attempted to make TECs accessible to everyone. They have succeeded, but in most cases (pun intended) the results will be no better than a normal air cooler and in many scenarios – worse. If you are serious about TECs, get a water cooling kit!