• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Dry Ice Cooling

No, it is usually only used for people wanting to try to maximize their overclock simply to get wcpuid scores so they can say see how high "I" got. It isn't for running all the time.

The dry ice method is usually this. A cup has its bottom cut out over the cpu core. Then it is siliconed on so it is liquid proof (notice I didn't say water). Then a mixture of alcohol (nearly pure) and dry ice (called a slurry I think) is poured in and then the computer is powered on. Usually these mobos have been voltage modded so that alot more voltage can be given and other mods have been performed.

Usually though instead of the dry ice slurry, people who go to this extreme push it one step further and break down and buy a small container of liquid nitrogen and use it instead.
 
Cooling with liquid nitrogen or (to a lesser extent) dry ice is pretty hard on a CPU.. The solder bumps of the FC package have to deal with a lot of stress from the different coefficients of thermal expansion of the silicon die and the OLPGA package. It doesn't take very many thermal cycles to have a fair amount of metal fatigue in the bumps and then to have breaks in the connections.
 
If you want better cooling than conventional air cooling then you should look into a watercooler. If you want extreme cooling (and money is not a problem) then I might suggest looking into a Vapochill (costs over $600 bucks).
 
Hey IMHO....I think that damaging these chips is 2 parts high voltage and 1 part high temps....I think that in most instances you are likley to be limited by cpu needing more voltage you can deliver or that is safe then temps holding you back...This chip will throttle itseldf as not to burn up.

Save your money and get a good 20-40 dollar air cooler max. I bet your 1.6 will be limited by voltage needs and stability long before temps are your limiting factor...
 
CMOS gets faster as it cools - the wires are less resistive and the mobility of the carriers in the transistors improves. Voltage changes get you more bang for the buck (in terms of performance improvement for percentage change), but if you cool a part down enough it will run pretty fast... if it works at all - if there are any marginal race conditions, then they will be exposed by the low temperature and the chip won't work at all. Since you can lower the temperature a lot more than you can increase the voltage (and still expect the part to work) extreme cooling is effective for increasing clock frequency substantially.

BUT thermally cycling at really high differentials is particularly hard on the package - and to a lesser extent the silicon crystal itself. The solder bumps of the FC package are subject to substantial shear stress and very quickly metal fatigue of the bumps can introduce reliability issues. In addition the crystalline structure of the silicon chip itself will start to see defects due to the thermal expansion coefficient differences between various layers, for example, the copper wiring and the silicon dioxide in the interconnect layers. Metal fatigure is likely to strike first and the chip may simply stop working suddenly after only a few thermal cycles (from ambient to LN temperatures), but even if metal fatigue doesn't kill the chip, the crystalline defects introduced will result in the chip running slower over the long term. Bottom line: I really wouldn't recommend doing this, but if you want to anyway, try to minimize the thermal shock to the chip.
 
Something you never see used but which should be very effective if somewhat noisy is a vortex tube chiller. If you have a source of compressed air you are good to go. Do a Google search and you'll seee what I mean. GoTo
 
One huge reason why you should avoid using dry ice is because if any water condenses near the ice, the CO2 will dissolve in that water making it fairly acidic. These acidic vapors will help corrode the metal parts inside of your case. That's one of the big reasons why liquid nitrogen is used to cool things off in most industries. Nitrogen is fairly inert and won't corrode things like carbon dioxide will.
 
😀 VERY interesting. It would be interesting to see one implemented on a CPU. Noise might be a big problem though (maybe the reason why it's not used😉).
 
Well, did a little research and I think the reason for Vortex coolers not being used is it's inherent low efficiency (only 4% of a heat pump).
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top