Mineral oil submerged rigs, more show than go to be honest,
oil isn't a very good heat conducter and once the heat is in the oil its trapped more or less so yes, more oil equals better (slightly) cooling capability but its still nowhere near waters performance
I've seen guides for fishtank setups but personally I wouldn't see it as cheap or easy
dug this out too for you, http://www.pugetsystems.com/submerged.php
What a mess! Can you imagine? What if you wanted to upgrade? How would gallons of warm oil smell in your house? This is one of those things that got talked about a lot a few years back, and some people did it, it does work, but it's one of those deals where the drawbacks outflank any of the possible benefits, I think anyway.
Another oil submersion idea...we have had a lot of these lately...
Messy, not as good as water, very long time to cool down and good luck selling used components after submersion- you basically need to soak in soapy water, rinse and allow to dry for a week before you can power back on. Oil also breaks down thermal paste, which isn't a good idea for any system.
Anyone thought about doing this with demineralized water and maybe an in-tank ion exchanger. We all know how awesome water is at moving heat. Maybe install some sort of heat pump to cool the water volume, and any cpu cooler with a large surface area would work well.
oh and FYI, you would need to clean every component COMPLETELY with demin water before installing and maintain the system in a clean environment (even with an ion exchanger). I would keep a conductivity meter in the water and make sure it stayed below .2 umho/cm^2 (that's really clean!). Any water lost to evaporation would need to be changed, and if your conductivity increased you would have to change the demineralizer resin. (and PLEASE make sure its H-OH resin and not like NH4-OH resin).
I would say run all fans in the case at a low constant speed, and have a variable speed fan to cool the entire water volume (to air) that doesn't kick on until the water gets to like 40*c. 10% speed at 40*, 50% @ 45*, 100% at 50*. That would keep any chip below its failure spec by a safe margin I think. If you were really OC'in a chip, maybe maintain those temps lower. Any thoughts?
I think either of these ideas aren't valid cooling options. Most of the concepts of heat output and thermal conduction/convection are being completely ignored. Oil does not absorb heat NEARLY as well as water- it takes a substantial amount of time to absorb the same amount of heat in watts. Likewise, it also take a long time to dissipate this heat. In the meantime, what about the excess heat from constantly running components? Where does it go? How do you plan to dissipate heat from a submerged system like this? Even if you decide to employ normal watercooling techniques of pump and radiator, remember that oil still doesn't transfer heat as quickly as water does.
Anyone thought about doing this with demineralized water and maybe an in-tank ion exchanger. We all know how awesome water is at moving heat. Maybe install some sort of heat pump to cool the water volume, and any cpu cooler with a large surface area would work well
So, you propose to immerse an entire PC into a tank of your solution and run expensive equipment designed to maintain pristine conditions within the tank? I don't agree that this is a viable solution...at all. You still haven't discussed how you plan to dissipate the heat being generated and absorbed by the water in the tank...simple dissipation due to surface area contact with the air will not be adequate for the amount of watts being constantly produced (but water/air surface contact will aid in keeping your de-ionizer busy with all the particulate matter being absorbed by the water). If you plan to seal it, I don't know what kind of heat pump you are discussing, but what I assume you are meaning is a water chiller, and these aren't that viable as 24/7 options when it comes to 700-800 watts of continual heat.