Cooling Considerations when Overclocking
One of the typical “true geek” hobbies is overclocking your computer. The basic idea is that you change several system settings, usually in BIOS, to cause your computer to run faster than it was designed to. This is an extremely dangerous process for your computer, and many considerations have to be taken when doing this. Not the least of which is how to cool your computer. Your standard fan speed, or your standard fan itself, isn’t going to cut it. Here’s an overview of what is being used to keep computers from burning up in these extreme conditions.
All electronic circuits create heat because of electron movement. As voltage applied and clock frequencies in the circuits increase, the heat created by components running at the increased clock speed increases. Because of this increased heat generated by overclocked components, more effective cooling is needed to keep from burning up the hardware. Also, some digital circuits wind down at high temperatures because of changes in the MOSFET device characteristics. Resistance in the wiring also increases at higher temperatures, adding up to decreased performance in the circuit.
Because most factory cooling systems are designed for non-overclocked use, overclockers usually turn to more advanced cooling systems, such as more powerful fans, larger heatsinks, water cooling and heat pipes. The best heatsinks are typically made completely of copper, which has high heat conductivity, but is more expensive. Aluminum is more widely used. While it has lower thermal conductivity, it is much cheaper than copper. Heat pipes are also used to improve conductivity.
Water cooling draws heat to a radiator. Thermoelectric cooling devices, aka Peltier devices, are now more popular with the onset of high Thermal Design Power (TDP) processors made by AMD and Intel. Thermoelectric cooling devices create temperature variances between two plates by running an electric current through them. This method is highly effective, but itself creates a lot of heat. Because of this, it is usually necessary to supplement these devices with a water-cooling system or a convection- based heatsink.
Other ways of pulling heat away from components are forced convection and phase change cooling. Liquid helium, liquid nitrogen, and dry ice are used in extreme cases, such as record-setting attempts or one-off experiments but not usually for an everyday system. These extreme methods are usually impractical in the long run, since they require refilling the reservoirs of vaporizing coolant, and also condensation can form on chilled components. Moreover, silicon-based components such as junction gate field-effect transistors (JFET) will lose effectiveness below temperatures of about 100 K (−173 °C; −280 °F) and will eventually freeze out at 40 K (−233 °C; −388 °F) since the silicon loses its semiconduction.
Submersion cooling, like what is used by the Cray-2 supercomputer, involves immersing a part of the computer system directly into a chilled fluid that conducts heat but has low electrical conductivity. The plus of this technique is that no condensation will form on components. A good submersion fluid is Fluorinert made by 3M, but it is expensive and can only be bought with a permit. Another option is mineral oil, but it might conduct electricity due to impurities.