Re: theory behind cooling
There are also less extreme forms of sub-ambient cooling such as Peltier and phase-change, either by having the cooling elements directly contact the CPU or by adding them to a watercooling loop. The benefits of sub-ambient cooling is something well-known to overclockers, but it's hard to find any direct tests to demonstrate it. Here's one I found, from '04: http://www.pcstats.com/articleview.c...id=1581&page=6
As you can see, even though the standard watercooling loop is doing a good job of dissipating the heat and keeping the CPU temperature down, the sub-ambient cooling is able to achieve a higher stable overclock.
Edit: Found exactly the sort of thing I was looking for! http://lab501.net/intel-core-i7-4790...king-study/26/
I think that's about as good a test as you could hope for, and with modern CPUs!