The fact that even while running a chiller, jpmboy has taken no condensation measures at all. This can only be done when his chilled water temps are above the dewpoint. Where the dewpoint is a mathematical function of ambient temperature and rellative humidity at a given time.
The dewpoint is NOT a fixed value, the ambient temp and the relative humidity changes even over a couple of hours. When running super chilled water below the dewpoint, I always calculate the dewpoint once every 2 hours or so.
Immediately if you get below the dewpoint, some serious actions have to be taken: prepping the motherboard, cpu socket and or gpu's identical to ln2 prepping: vasi, plasti spray, eraser electrical tape and last but not least any part from the cooling loop that is exposed to ambient temps will condensate like crazy: everything has to be packed In armaflex and completely sealed of or water will be pooring over your pressious hardware prior to it making its final sudden shutdown resulting in a pile of dead junk.
I run a Haileah 1500watt chiller, 1 noisy MF, I have put insulated copper pipping fixed from the wall where my bench desk resides. Al the way through my kitchen, through a wall into another room.. THe chiller is sitting there (silence I Cill you) placed the chiller about 25 meter from my desk, so about so a loop jength 50-60 meter back and forth. Needs 2 strongest Koolance PMP500 pumps to get a flow that sits in 300-400 liters per hour, pressure is 0,7-0,8 bar
Mine drops to 4 `C, so on average that is an immediate bonus from 14 to 24 `C on the water temp. On top of that when de system is stressed the water temps rises only by a fractions simply because the chiller is way fast in resupply of freshly chilled water.
CPU cores idle at 8`C. Depending on test and how hard it's pushed underload max around 85 sofar.
But GPU .... They loveeeee the cold imagine core at 11`C and vrms 17-19 `C . Max under load 35`c!! And they love that