Slashdot Mirror
How to Cool Your PC with Dry Ice
Ant writes "This Madshrimps article is a complete guide to working with dry ice so you can reach sub-zero temperatures with your CPU and graphics card. Details on building containers, where to buy dry ice and important tips and tricks. (Seen on Blue's News.)"
When has a hack or a weird project like this ever rendered a logical concise answer to the question "Why?" other than "Just because" or "To see if it could be done"?
On the plus side, this device hardly seems like something that could be mass-produced profitably. On the minus side, the dangers of dry ice, the border-line exorbitant cost of maintenance, and what appears to be a customized fit to the PC at hand (i.e.: it doesn't look like they could take that monster out and put it into another PC) makes me wonder what the point of this is. And I think I know exactly ;)
It would get rather expensive using dry ice over time, somehow I think its more a "proof-of-concept" rather than useful in the real world.
http://www.fanboy.co.nz/adblock/
Placing granular dry-ice into a copper container next to your CPU/GPU isn't going to result in very efficient heat transfer. If you're really going to try this, I recommend floating the dry ice in some sort of organic solvent (don't try water, as it'll freeze) and setting up some sort of agitator system.
An even better setup would be to setup a water-cooling-like system, with dry ice in a container filled with solvent above the system. The solvent would be fed into a heat exchanger on top of the CPU/GPU. No pump would be necessary, because, as the solvent in the heat exchanger heats up, it'll float up to the top, where it'll be cooled down again by the dry ice.
In any case, the costs of such a system would probably outweigh any benefit of using it.
I'm Trappped at Berkeley.
Sorry, but the cells will generate enough heat themselves, that no fan would be able to remove it all. Actually, in stack of 4-5 cells, the but-last (before the one sticking to the cooler) would likely explode, boiling. One, well adjusted thermoelectric cell, with really good cooling (water, really big fan) can be useful and actually decrease the CPU power. Adding the second one will already produce more heat than any commonly available solution (fans, water cooling) can remove. Maybe dry ice could remove the surplus heat, but then, why not use the dry ice straight on the radiator? It would be more efficient.
Anagram("United States of America") == "Dine out, taste a Mac, fries"
Rime from water and other substances is a major problem in cryo work. So is heat transfer. We normally use acetone or isopropyl alcohol in vacumm trap cold bath dewars to improve heat transfer.
And when you get tired of overclocking, put a handful of magnesium turnings in the center of a big block of dry ice. Ignite.
Note that the resulting fire is not extinguished by either water or CO2. It can also blind you or permanently damage your retina and/or burn you to death.
But what the hey, it's fun!
Because practical solutions to problems interest me. Silly, not-practical stuff like this does not.
Do what you want to. I think experimentation fine. Get back to me when it's usable.
Why yes, I AM a rocket scientist!