Computer cooling is required to remove the waste heat generated by computer components, to keep components within the permissible operating temperature limits. Components susceptible to temporary malfunctions or permanent damage if overheating include integrated circuits such as Central Processing Units (CPUs), chipsets, graphics cards, and hard disk drives.
Components are often designed to produce as little heat as possible, and computers and operating systems can be designed to reduce power consumption and heating due to workload, but more heat can still be produced than can be removed without regard to cooling. The use of heatsinks cooled by airflow reduces the temperature rise generated by the amount of heat provided. Attention to airflow patterns may prevent hotspot development. Computer fans are widely used along with heatsink fans to reduce temperature by actively exhausting hot air. There are also more exotic cooling techniques, such as liquid cooling. All modern processors are designed to cut or reduce the voltage or clock speed if the internal temperature of the processor exceeds the specified limit.
Refrigeration may be designed to reduce ambient temperature in computer cases, such as exhausting hot air, or cooling one component or small area (cooling point). Individually cooled components include CPU, graphics processing unit (GPU) and northbridge.
Video Computer cooling
Generator of unwanted heat
Integrated circuits (eg, CPU and GPU) are the main heat generators in modern computers. Heat generation can be reduced by efficient design and selection of operating parameters such as voltage and frequency, but ultimately, acceptable performance can often only be achieved by managing significant heat.
In operation, the temperature of the computer component will rise until the heat transferred to the surrounding equals the heat generated by the component, ie when thermal equilibrium is reached. For reliable operation, the temperature should not exceed the maximum maximum allowed for each component. For semiconductors, the instantaneous connection temperature, rather than component case, heatsink, or ambient temperature is very important.
Refrigeration can be changed by:
- Dust acts as a thermal insulator and blocks airflow, reducing heat sink and fan performance.
- Poor airflow including turbulence due to friction against barrier components such as ribbon cable, or faulty fan orientation, can reduce the amount of air flowing through the casing and even create a localized air vortex in hot air at case. In some cases of apparatus with poor thermal design, cooling air can easily flow out through "cooling" the holes before passing through the heat component; cooling in such cases can often be enhanced by blocking the selected holes.
- Bad heat transfer due to poor thermal contact between the component to be cooled and the cooling device. This can be increased by using a thermal compound to remove surface imperfection, or even by lapping.
Maps Computer cooling
Prevention of damage
Because high temperatures can significantly reduce the lifetime or cause permanent damage to components, and the heat output from components can sometimes exceed the computer's cooling capacity, manufacturers often take additional precautions to ensure that temperatures remain within safe limits. Computers with thermal sensors integrated in the CPU, motherboard, chipset, or GPU can shut down when high temperatures are detected to prevent permanent damage, although this may not fully guarantee long-term safe operation. Before the overheating component reaches this point, it may "strangle" until the temperature falls below the safe point using dynamic frequency scaling technology. Refinements reduce the operating frequency and voltage of integrated circuits or disable non-essential features of the chip to reduce heat output, often with little or significantly reduced performance cost. For desktop and notebook computers, throttling is often controlled at the BIOS level. Throttling is also commonly used to regulate the temperature in smartphones and tablets, where components are packed tightly together with little to no active cooling, and with additional heat transferred from the user's hands.
Mainframe and supercomputer
As electronic computers become larger and more complex, cooling of active components becomes an important factor for reliable operation. Early vacuum-tube computers, with relatively large cabinets, can rely on natural or forced air circulation for cooling. However, solid state devices are packed denser and have lower operating temperatures allowed.
Beginning in 1965, IBM and other mainframe computer manufacturers sponsored intensive research into the solid, integrated physics of circuits. Many air-conditioning and liquid systems are designed and investigated, using methods such as natural and forced convection, direct air strikes, direct fluid immersion and forced convection, boiling pools, falling films, boiling streams, and liquid jet slippage. Mathematical analysis is used to predict the temperature rise of components for any possible geometry of the cooling system.
IBM developed three generations of Thermal Conduction Module (TCM) that uses a water-cooled cooling plate in direct thermal contact with integrated circuit packs. Each package has a thermally conductive pin pressed onto it, and a helium gas chamber is surrounded and hot regulating pins. This design can remove up to 27 watts from the chip and up to 2000 watts per module, while maintaining the chip package temperature of around 50 Â ° C (122 Â ° F). Systems using TCM were 3081 families (1980), ES/3090 (1984) and some models ES/9000 (1990). In the IBM 3081 processor, TCM allows up to 2,700 watts on a single printed circuit board while maintaining chip temperature at 69 ° C (156 ° F). Thermal conduction modules using water cooling are also used in mainframe systems manufactured by other companies including Mitsubishi and Fujitsu.
Cray-1 supercomputer designed in 1976 has a distinctive cooling system. The machine is only 77 inches (2,000 mm) and 56 1 / 2 inches (1,440 Ã, mm) with diameter, and consumed up to 115 kilowatts; this is proportional to the average power consumption of several dozen Western homes or medium-sized cars. The integrated circuit used in the machine was the fastest of the time, using logic paired with the emitter; However, the speed is accompanied by high power consumption compared to newer CMOS devices.
Heat removal is very important. Refrigerant is circulated through piping which is embedded in a vertical cooling bar in twelve parts of columnar machines. Each of the 1662 printed circuit modules from the machine has a copper core and is clamped to a cooling bar. The system is designed to keep the case of integrated circuits at no more than 54 Ã, Â ° C (129Ã, Â ° F), with the refrigerant circulating at 21Ã, Â ° C (70Ã, Â ° F). The final heat rejection is through a water-cooled condenser. Pipes, heat exchangers, and pumps for cooling systems are arranged in bench seats around the outside of the computer base. Approximately 20 percent of the operating weight of the machine is refrigerant.
In the later Cray-2, with more dense modules, Seymour Cray has the problem of effectively cooling the engine using metal conduction techniques with mechanical cooling, so it switches to cooling 'liquid immersion'. This method involves charging a Cray-2 chassis with a liquid called Fluorinert. Fluorinert, as the name implies, is an inert liquid which does not interfere with the operation of electronic components. When the component comes to the operating temperature, the heat will dissipate into Fluorinert, which is pumped out of the machine into a cold water heat exchanger.
Performance per watt of modern systems has been greatly improved; more calculations can be made with given power consumption than is possible with integrated circuits of the 1980s and 1990s. Recent supercomputer projects such as Blue Gene rely on air cooling, which reduces the cost, complexity, and size of the system compared to liquid cooling.
Air cooling
Fans
The fan is used when natural convection is not enough to remove heat. The fan may be mounted on the computer case or attached to the CPU, GPU, chipset, PSU, hard drive, or when the card is plugged into the expansion slot. Common fan sizes include 40, 60, 80, 92, 120, and 140 mm. Fans 200, 230, 250, and 300 mm are sometimes used in high-performance personal computers.
Chassis fan performance
The computer has a certain resistance to airflow through the chassis and components. This is the sum of all the smaller obstacles to airflow, such as inlet and outlet openings, air filters, internal chassis, and electronic components. The fan is a simple air pump that provides pressure to the inlet air relative to the output side. The pressure difference is moving air through the chassis, with air flowing into the lower pressure area.
Fans generally have two published specifications: free air flow and maximum differential pressure. Free air stream is the amount of fan air will move with zero back pressure. The maximum differential pressure is the amount of pressure a fan can generate when it is completely blocked. Among these two extremes is a series of corresponding measurements of the flow versus the pressure usually presented as a graph. Each fan model will have a unique curve, such as a dashed curve in adjacent illustrations.
Installing parallel versus series
Fan can be installed parallel to each other, in series, or a combination of both. The parallel installation is a fan installed side by side. The series installation will be a second fan in accordance with other fans such as inlet fans and exhaust fans. To simplify the discussion, it is assumed the fans are the same model.
Parallel fans will provide twice as much free air flow but no additional driving pressure. The series installation, on the other hand, will double the available static pressure but does not increase the free airflow rate. The adjacent illustration shows a fan versus two fans in parallel with a maximum pressure of 0.15 inches (3.8 mm) of water and a flow rate doubling about 72 cubic feet per minute (2.0 m 3 /mnt).
Note that the airflow changes as the square root of the pressure. Thus, doubling pressure will only increase the flow of 1.41 (span>? 2 ) times, not double that assumed. Another way to look at this is that the pressure should go up by a factor of four to double the flow rate.
To determine the flow rate through the chassis, the chassis impedance curve can be measured by imposing indiscriminate pressure on the inlet to the chassis and measuring the flow through the chassis. This requires quite sophisticated equipment. With the chassis impedance curve (represented by solid red and black lines on adjacent curves) is determined, the actual flow through the chassis as generated by certain fan configurations is shown graphically where the chassis impedance curve crosses the fan curve. The slope of the chassis impedance curve is a square root function, where doubling the flow rate is four times the differential pressure.
In this particular example, adding a second fan provides a marginal improvement with flow for both configurations to about 27-28 cubic feet per minute (0.76-0.79 m 3 /min). Although not shown on the plot, the second fan in series will provide slightly better performance than parallel installs.
Temperature versus flow rate
The equation for the airflow required through the chassis is
where
CFM = Cubic Feet per Minute (0,028 m 3 /min) Q = Heat Transferred (kW) Cp = Air Specific Heat r = Density DT = Temperature Change (in  ° F)
A simple conservative rule for the needs of cooling flow, discounting effects such as heat loss through chassis walls and laminar versus turbulent flow, and accounting for constants for Specific Heat and Density at sea level are: (Please Note That must be between sea levels)
For example, a typical chassis with a 500 watt load, maximum air temperature of 130 Â ° C (54 Â ° C) in a 100 Â ° F (38 Â ° C) environment, ie a 30 Â ° F (17 Â ° C) difference. Â ° C):
This will be the actual flow through the chassis and not the fan-free air rating.
Piezoelectric pump
"Dual piezo cooling jet", patented by GE, uses vibration to pump air through the device. The initial device is three millimeters thick and consists of two nickel plates connected on both sides to a piece of piezoelectric ceramic. The current back and forth through the ceramic component causes it to expand and contract up to 150 times per second so that the nickel disc acts like a bellows. Contracted, the tip of the disc is pushed together and sucks the hot air. Expands to carry nickel discs together, throw air at high speed.
The device has no pads and does not require a motor. It's thinner and consumes less energy than regular fans. Jet can move the same amount of air as the cooling fan twice the size while consuming half the electricity and the lower cost.
Passive cooling
Passive heat-sink heaters involve sticking machine blocks or metal extrusions into parts that require cooling. Thermal adhesive can be used. More commonly for personal computer CPUs, clamps hold heat sinks directly over chips, with thermal grease or thermal pad spread out between the two. This block has fins and ridges to increase its surface area. The thermal conductivity of the metal is much better than air, and it emits heat better than a protected component (usually an integrated circuit or CPU). Heat sink cooling-cooling fans initially are the norm for desktop computers, but today many heat sinks have copper base plates or are made entirely of copper.
The buildup of dust between the metal fins of the heat sink gradually reduces the efficiency, but can be recovered with a gas towel by blowing dust along with other unwanted excess materials.
Passive heat sinks are commonly found on older CPUs, less hot parts (such as chipsets), and low-power computers.
Usually a heat sink is attached to an integrated heat spreader (IHS), essentially a large, flat plate attached to the CPU, with a conduction paste lined between. It removes or distributes heat locally. Unlike heat sinks, dispersers are meant to distribute heat, not to eliminate it. In addition, IHS protects the fragile CPU.
Passive cooling does not involve fan noise as the force of convection moves air above the heatsink.
Other techniques
Liquid immersion cooling
Unusual practice is to immerse computer components in hot liquids, but not electrically. Although rarely used for personal computer cooling, liquid immersion is a routine method of cooling large power distribution components such as transformers. It also became popular with data centers. Personal computers that are cooled in this way may not require fans or pumps, and can be cooled exclusively by the passive heat exchanges between the computer hardware and the plastic boxes placed therein. The heat exchanger (ie the heater or radiator core) may still be needed though, and the piping also needs to be placed properly.
The liquid used must have a low enough electrical conductivity to not interfere with the normal operation of the computer. If the liquid is electrically conductive, it may cause electrical shorts between components or traces and permanently damage them. For this reason, it is preferred that the liquid be an insulator and not conduct electricity..
A wide variety of liquids are available for this purpose, the most suitable transformer oil and special cooling oils such as 3M Fluorinert. Non-purpose oils, including cooking oil, motor and silicon, have been successfully used to cool personal computers.
Evaporation can cause problems, and fluids may need to be refilled regularly or sealed in a computer enclosure.
Reduce heat loss
Where powerful computers with many features are not needed, less powerful computers or those with fewer features can be used. In 2011 VIA EPIA motherboards with CPUs typically eliminate about 25 watts of heat, while more capable Pentium 4 motherboards and CPUs typically disappear about 140 watts. Computers can be enabled with direct current from external power bricks that do not generate heat inside the computer case. The replacement of the cathode ray tube (CRT) displayed by a more efficient liquid crystal display screen (LCD) in the early twenty-first century significantly reduces power consumption.
Heat-sinks
A component can be mounted in a good thermal contact with a heatsink, a passive device with a large heat capacity and with a large surface area relative to its volume. Heatsinks are usually made of metal with high thermal conductivity such as aluminum or copper, and incorporate fins to increase the surface area. Heat from a relatively small component is transferred to a larger heatsink; the equilibrium temperature of the component plus the heatsink is much lower than the component alone. The heat is carried from the heatsink by convective or fan-forced airflow. Fan cooling is often used to cool processors and graphics cards that consume large amounts of electrical energy. On computers, common heat-generating components can be made with flat surfaces. A metal beam with a suitable flat surface and finned construction, sometimes with an installed fan, clamped onto a component. For bad filling do the air gap because the surface is not perfect and smooth, thin layer of thermal grease, thermal pad, or thermal adhesive can be placed between the component and heatsink.
Heat is removed from heat-sink by convection, to some extent by radiation, and possibly by conduction if the heat-sink is in thermal contact with, say, a metal casing. The inexpensive cooling aluminum-fan coolers are often used on standard desktop computers. Heat-sinks with copper base plates, or made of copper, have better heat characteristics than aluminum. Copper heats sinks are more effective than aluminum units of the same size, which are relevant to the high power consumption components used in high-performance computers.
Passive heat sinks are commonly found on: older CPUs, parts that do not remove much power, such as chipsets, computers with low-power processors, and equipment where silent operation is essential and fan noise is unacceptable.
Usually heat-sinks are clamped onto an integrated heat spreader (IHS), flat metal plate size of the CPU package that is part of the CPU assembly and spreads heat locally. Thin layer of thermal compound is placed between them to compensate for surface imperfection. The main purpose of spreaders is to distribute heat. The heat sink fins enhance their efficiency.
Some brands of DDR2, DDR3, DDR4 and upcoming DDR5 DRAM memory modules are equipped with finned heatsinks captured to the top edges of the module. The same technique is used for video cards that use passive heatsink fining in the GPU.
Dust tends to accumulate in the cracks of finned heatsinks, especially with the high airflow produced by fans. It removes air from heat components, reduces cooling effectiveness; However, removing dust restores the effectiveness.
Peltier (thermoelectric) cooling
Peltier joints are generally only about 10-15% as efficient as the ideal refrigerator (Carnot cycle), compared to the 40-60% achieved by conventional compression cycle systems (reverse Rankine systems using compression/expansion). Because of this lower efficiency, thermoelectric cooling is generally used only in environments where solid state properties (no moving parts, low maintenance, compact size, and insensitive orientation) exceed pure efficiency.
Modern TEC uses several stacked units each consisting of dozens or hundreds of thermocouples placed next to each other, allowing for a large amount of heat transfer. The combination of bismuth and tellurium is most often used for thermocouples.
As an active heat pump that consumes power, the TEC can produce temperatures below ambient, impossible with passive heatsinks, radiator cooled cooling chillers, and heatpipe HSFs. However, when pumping heat, the Peltier module will typically consume more electrical power than the amount of heat pumped.
Liquid cooling
Liquid cooling is a very effective method of removing excess heat, with the most common heat transfer fluid in desktop PCs being distilled water. The advantages of cooling water through air cooling include higher specific water heat capacity and thermal conductivity.
The principle used in a typical liquid cooling system (active) for a computer is identical to that used in a car's internal combustion engine, with water circulated by a water pump through waterblock mounted on the CPU (and sometimes additional components as GPU and northbridge) and out to a heat exchanger, usually a radiator. The radiator itself is usually cooled extra by using a fan. In addition to fans, it may also be cooled in other ways, such as Peltier cooling (although the Peltier element is most often placed directly above the hardware to cool, and the coolant is used to heat from the thermal side of the Peltier element). Cooling reservoirs are often also connected to the system.
In addition to the active liquid cooling system, passive liquid cooling systems are also sometimes used. These systems often dispose of fans or water pumps, thereby theoretically increasing the reliability of the system, and/or making it more quiet than the active system. The downside of this system is that they are less efficient at wasting heat and thus also need to have more cooling - and thus a much larger cooling reservoir - (giving more time for cooling to cool down).
The liquid allows more heat transfer from the cooled parts of the air, making liquid cooling suitable for overclocking and high performance computer applications. Compared with air conditioning, fluid cooling is also affected less by ambient temperature. Liquid cooling noise levels are relatively low compared to active cooling, which can be very noisy.
Disadvantages of coolant fluids include the complexity and potential leakage of coolant fluids. Water leaking (or more importantly additives added to water) can damage electronic components that come into contact with it, and the need to test and repair leaks makes installations more complex and less reliable. (Especially, the first major attack into the field of liquid-cooled personal computers for general use, high-end versions of the Apple Power Mac G5, ultimately destined by a tendency to leak coolers.) An air-cooled heat sink is generally easier to build, and maintain rather than water cooling solutions, although special CPU water cooling kits can also be found, which may be just as easy to install as air conditioning. It's not limited to CPUs, however, and liquid cooling of GPU cards is also possible.
Although initially confined to mainframe computers, liquid cooling has become a practice largely associated with overclocking in the form of manufactured kits, or in the form of do-it-yourself arrangements that are assembled from individually collected parts. In recent years there has been an increase in the popularity of fluid cooling in desktop computers that have been assembled before, medium to high. A closed-loop system incorporating a pre-filled radiator, fan, and small waterblock simplifies installation and maintenance of cooling water at little cost in cooling effectiveness relative to larger and more complex settings. Liquid cooling is usually combined with air cooling, using liquid cooling for the hottest components, such as a CPU or GPU, while retaining a simpler and cheaper air cooling for less demanding components.
The Aquasar System of IBM uses hot water coolers to achieve energy efficiency, water used to heat buildings as well.
Since 2011, the effectiveness of water cooling has led to a series of all-in-one water cooling solutions (AIOs). The AIO solution produces a simpler mounting unit, and most units have been reviewed positively from the review site.
Heat pipe
The heat pipe is a vacuum tube filled with heat transfer fluids. The liquid absorbs heat and evaporates at one end of the pipe. Steam flows to the other end of the tube (cold), where it condenses, releases its latent heat. The liquid returns to the hot end of the tube by gravity or capillary and repeats the cycle. The heat pipe has an effective heat conductivity much higher than the solid material. For use in computers, the heat sink on the CPU is attached to the larger radiator heat sink. Both heat sinks are void, such as the attachment between them, creating a large heat pipe that transfers heat from the CPU to the radiator, which is then cooled using some conventional methods. This method is expensive and is usually used when the space is narrow, as in PCs and small-form factor PCs, or where no fan sounds can be tolerated, as in audio production. Due to the efficiency of this cooling method, many desktop CPU and GPUs, as well as high end chipsets, use hot pipes in addition to active fan-based cooling to stay in safe operating temperatures.
Electrostatic air movement and cooling corona release effect
The cooling technology being developed by Kronos and Thorn Micro Technologies uses a device called an ionic wind pump (also known as an electrostatic fluid accelerator). The basic operating principle of the ionic wind pump is the release of the corona, the discharge of electricity near the charged conductor caused by the ionization of the surrounding air.
The corona wastewater cooler developed by Kronos works in the following ways: A high electric field is created at the end of the cathode, which is placed on one side of the CPU. High energy potentials cause the oxygen and nitrogen molecules in the air to become ionized (positively charged) and create a corona (charged particle circle). Placing the anode ground at the end of the CPU causes the charged ion in the corona to accelerate toward the anode, colliding with the neutral air molecules on the road. During this collision, momentum is transferred from the ionized gas to the neutral air molecule, producing a gas movement toward the anode.
The advantage of a corona-based coolant is the lack of moving parts, thus eliminating certain reliability problems and operating with near-zero noise levels and moderate energy consumption.
Soft cooling
Soft cooling is the practice of utilizing software to take advantage of CPU power saving technologies to minimize energy use. This is done by using the termination instruction to shut down or insert in standby state CPU part that is not used or with underclocking CPU. While generating a lower total speed, this can be very useful if overclocking the CPU to improve the user experience rather than increasing the processing power of the crud, as it can prevent the noisy cooling needs. Contrary to what the term suggests, this is not a form of cooling but reduces heat creation.
Undervolting
Undervolting is the practice of running a CPU or other component with a voltage below the device specification. The non-attenuated component draws less power and thus produces less heat. The ability to do this varies by manufacturer, product line, and even different production runs of the same product (as well as other components in the system), but the processor is often determined to use higher voltages than is necessary. This tolerance ensures that the processor will have a higher chance of performing properly under sub-optimal conditions, such as low-quality motherboards or low power supply voltages. Below a certain limit, the processor will not work correctly, although undervolting too far usually does not lead to permanent hardware damage (unlike overvolting).
Undervolting is used for quiet systems, as less cooling is required due to reduced heat production, allowing noisy fans to be removed. This is also used when the battery power must be maximized.
Chip-integrated
Conventional cooling techniques all attach their "cooling" components to the outside of the computer chip package. This "sticky" technique will always show some thermal barriers, reducing its effectiveness. Heat can be more efficiently and quickly removed by directly cooling the local hot spot of the chip, inside the package. At this location, power dissipation of more than 300 W/cm 2 (typical CPUs less than 100 W/cm 2 ) may occur, although the future system is expected to exceed 1000 W/cm 2 . This form of local cooling is important to develop chips with high power density. This ideology has led to the investigation of integrating cooling elements into computer chips. Currently there are two techniques: micro channel heat sink, and jet jet refrigeration.
In a micro channel heat sink, channels are made into the silicon chip (CPU), and the coolant is pumped through it. Channels are designed with a very large surface area that produces large heat transfer. Heat dissipation 3000 W/cm 2 has been reported with this technique. Heat dissipation can be further improved if two phase phase cooling is applied. Unfortunately, this system requires great pressure, due to small channels, and lower heat flux with dielectric cooling used in electronic cooling.
Another local chip cooling technique is jet jet cooling. In this technique, the coolant is flown through a small hole to form a jet. Jets are directed to the CPU chip surface, and can effectively remove large heat fluxes. Heat dissipation over 1000 W/cm 2 has been reported. This system can be operated at low pressure compared to micro channel method. The heat transfer can be increased further by cooling the two-phase flow and by integrating the reverse channel (hybrid between the micro channel heat sink and jet jet refrigeration).
Phase-change cooling
Cooling phase change is a very effective way to cool the processor. The phase-change vapor refrigeration compressor is a unit that usually sits beneath the PC, with a tube leading to the processor. Inside the unit is a compressor of the same type as in AC. The compressor presses the gas (or gas mixture) into the liquid. Then, the liquid is pumped into the processor, where it passes through the condenser (heat exchanger) and then the expansion means to vaporize the liquid; the expansion tool used can be a simple capillary tube to a more complex thermal expansion valve. The liquid evaporates (phase change), absorbs heat from the processor as it draws extra energy from its environment to accommodate this change (see latent heat). Evaporation can produce temperatures of about -15 to -150 ° C (5 to -238 ° F). The gas flows into the compressor and the cycle begins again. In this way, the processor can be cooled to temperatures ranging from -15 to -150 ° C (5 to -238 ° F), depending on the load, wattage of the processor, cooling system (see cooling) and mixed gas used. This type of system suffers from a number of problems (cost, weight, size, vibration, maintenance, electricity costs, noise, the need for special computer towers) but, in particular, one should be concerned with the dew point and proper insulation of all sub-ambient surfaces done (pipe will sweat, dripping water on sensitive electronics).
Alternately, a new generation of cooling systems is being developed, incorporating the pump into the thermosiphone loop. This adds another level of flexibility to the design engineer, since heat can now be effectively transported away from heat sources and reclaimed or thrown into the ambient. The junction temperature can be adjusted by adjusting the system pressure; a higher pressure equals a higher liquid saturation temperature. This allows smaller condensers, smaller fans, and/or effective heat dissipation in high environmental temperature environments. This system is, in effect, the next generation liquid cooling paradigm, since they are approximately 10 times more efficient than single-phase water. Because this system uses dielectrics as a heat transport medium, leaks do not cause electrical system failures.
This type of cooling is seen as a more extreme way to cool the components, because the units are relatively expensive compared to the average desktop. They also produce a large amount of noise, since they are essentially fridges; however, the choice of compressors and air conditioning systems is a key determinant of this, allowing flexibility for noise reduction based on selected sections.
Liquid nitrogen
As liquid nitrogen boils at -196Ã, Â ° C (-320,8Ã, Â ° F), well below the freezing point of water, it is valuable as an extreme cooler for short overclocking sessions.
In special installations of liquid nitrogen cooling, copper or aluminum pipes are mounted on top of the processor or graphics card. Once the system has been very isolated against condensation, liquid nitrogen is poured into the pipe, resulting in a temperature below -100 ° C (-148 ° F).
Evaporation devices range from cutting off heat sinks with pipes attached to special milled copper containers used to hold nitrogen as well as to prevent large temperature changes. However, once the nitrogen evaporates, it must be refilled. In the field of personal computers, this cooling method is rarely used in contexts other than experimental overclocking and record-keeping attempts, since the CPU will typically expire in a relatively short time due to temperature stress caused by changes in internal temperature.
Although liquid nitrogen is not flammable, it can condense oxygen directly from the air. Liquid oxygen mixtures and flammable substances can be dangerous explosives.
Liquid nitrogen cooling, in general, is only used for processor benchmarking, due to the fact that continuous use can cause permanent damage to one or more computer parts and, if handled in a reckless manner, may even harm the user.
Liquid helium
Liquid helium, colder than liquid nitrogen, has also been used for cooling. Liquid helium boils at -269 Â ° C (-452.20 Â ° F), and temperatures ranging from -230 to -240 Â ° C (-382.0 to -400.0 Â ° F) have been measured from the heatsink. However, liquid helium is more expensive and more difficult to store and use than liquid nitrogen. Also, very low temperatures can cause integrated circuits to stop functioning. Silicon-based semiconductors, for example, will freeze about -233 Â ° C (-387.4 Â ° F).
Optimization
Cooling can be improved by some techniques that may involve additional costs or effort. These techniques are often used, in particular, by those who run their computer parts (such as CPU and GPU) at higher voltages and frequencies than those specified by the manufacturer (overclocking), which increases heat generation.
Installation of higher performance, non-stock cooling can also be considered modding. Many overclockers only purchase more efficient, and often more expensive, fan combinations and heat sinks while others use more exotic computer cooling methods, such as fluid cooling, pelvic heatpump effects, heat pipe or phase change cooling.
There are also some related practices that have a positive impact on reducing system temperature:
Thermal conductive compound
Often called Thermal Interface Materials (TIMs) (eg Intel)
The perfectly flat surface in contact provides optimal cooling, but perfect flatness and absence of microscopic air gaps are not practically possible, especially in mass-produced equipment. A very thin thermal compound, which is much more thermally conductive than air, although much less than metal, can increase thermal and cooling contacts by filling air gaps. If only a small amount of compound is sufficient to fill the gap used, the best temperature reduction will be obtained.
There is much debate about the benefits of compounds, and overclockers often assume some compounds are superior to others. The main consideration is to use the minimum amount of thermal compound required for a more even surface, since the thermal conductivity of the compound is usually 1/3 to 1/400 compared to the metal, although much better than air. Conductivity of heatsink compounds ranges from about 0.5 to 80W/mK (see article); that of aluminum about 200, that air is about 0.02. Heat-conductive pads are also used, often installed by manufacturers for heatsinks. They are less effective than correct applied thermal compounds, but are easier to apply and, if fixed to a heatsink, can not be eliminated by the user not realizing the importance of good thermal contact, or being replaced by a thick and ineffective layer of compounds.
Unlike some of the techniques discussed here, the use of thermal or padding compounds is almost universal when scattering large amounts of heat.
Heat whapping
The spread of CPU heat is mass-produced and the heatsink base is never really flat or smooth; if this surface is placed in the best contact, there will be an air gap that reduces heat conduction. This can be easily reduced by the use of thermal compounds, but for best results, the surface should be as uniform as possible. This can be achieved by a grueling process known as lapping, which can reduce the CPU temperature by typically 2 Â ° C (4 Â ° F).
Round cable
Most older PCs use flat ribbon cables to connect storage drives (IDE or SCSI). This large flat cable greatly inhibits airflow by causing drag and turbulence. Overclockers and modders often replace this with rounded cables, with conductive cables held together to reduce surface area. Theoretically, the parallel thread of the conductor in the ribbon cable serves to reduce the crosstalk (a signal carrying the conductor induces the signal at the nearest conductor), but there is no empirical evidence of the rounding cable that reduces performance. This may be because the cable length is short enough so that the crosstalk effect can be ignored. Problems usually arise when the cable is not electromagnetically protected and the length is large enough, more common with older network cables.
These computer cables can then become cables that are fastened to the chassis or other cables to further increase airflow.
This is less of a problem with newer computers that use Serial ATA that have much narrower cables.
Airflow
The cooler the cooling medium (air), the more effective it is cooling. Cooling air temperature can be improved with this guide:
- Fill the cool air into the heat component as soon as possible. An example is a snorkel and an air tunnel that supplies the outside air directly and exclusively to the CPU or GPU cooler. For example, the BTX case design determines the CPU air tunnel.
- Wash warm air as soon as possible. Examples are: Conventional PC power supply (ATX) blows warm air out of the back of the case. Many designs of dual-slot graphics cards blow warm air through adjacent slot covers. There are also some aftermarket coolers that do this. Some cooling CPU designs blow warm air directly into the back of the casing, where it can be ejected by case fans.
- The air that has been used to cool a component should not be reused to cool different components (this follows from the previous item). The BTX case design violates this rule, as it uses the CPU cooling exhaust to cool the chipset and often the graphics card. One can find old or ultra-low ATX cases featuring PSU mounting at the top. Most modern ATX cases do have a PSU mounted on the bottom of the casing with air vent being filtered directly under the PSU.
- Prefers cold air, avoid breathing exhaust air (outside air above or near exhaust). For example, the CPU cooler channel at the back of the tower box will suck the warm air from the graphics card's exhaust. Moving all discharges to one side of the chassis, conventionally the back/top, helps keep the air cool.
- Hiding the cables behind the motherboard tray or simply use a ziptie and tuck a cable to provide uninterrupted airflow.
Fewer but strategically placed fans will increase internal airflow within the PC and thereby lower the overall internal temperature of the case in relation to ambient conditions. The greater use of fans also increases efficiency and decreases the amount of waste heat along with the amount of noise generated by the fans while operating.
There is little agreement on the effectiveness of different fan placement configurations, and few ways of systematic testing have been done. For rectangular PC case (ATX), front fan with fan on the back and one at the top has been found as the appropriate configuration. However, AMD's (slightly outdated) cooling system guideline notes that "The front cooling fan does not seem to be important.In fact, in some extreme situations, testing shows this fan will drain hot air instead of using cold air." Perhaps fans on the side panel could have the same detrimental effect - perhaps by interrupting normal airflow through the casing. However, this has not been confirmed and may vary with configuration.
Computer type
Desktop
Desktop computers typically use one or more fans for cooling. While almost all desktop power supplies have at least one built-in fan, the power supply should not draw hot air from inside the casing, as this results in higher PSU operating temperatures that decrease the energy efficiency, reliability, and overall ability of the PSU to provide power supply to the internal components of the computer. For this reason, all modern ATX cases (with some exceptions found in very low budget cases) feature a power supply at the bottom, with a special air intake of the PSU (often with its own filter) below the mounting location, allowing the PSU to draw cold air from under the casing.
Most manufacturers recommend bringing cool, fresh air at the bottom of the front of the chassis, and exhausting the warm air from the top of the rear. If the fan is installed to force air into the casing more effectively than removed, the inside pressure becomes higher than the outside, which is called the "positive" air flow (opposite case is called "negative" airflow). It should be noted is that positive internal pressure only prevents the accumulation of dust in this case if the air intakes are equipped with a dust filter. A case with negative internal pressure will experience a higher level of dust accumulation even if the intake is filtered, since negative pressure will attract incoming dust through the available opening in this case.
Airflow inside a typical desktop case is usually not strong enough for a passive CPU heatsink. Most desktop heat sinks are active including one or even some directly installed fan or blower.
Server
The server cooling fan in the slot (1 U) is usually in the middle of the enclosures, between the front hard drive and the passive CPU cooler on the back. Larger (higher) scopes also have a fan, and from about 4U they may have an active heat sink. The power supply generally has its own back-facing exhaust fan.
Installed on the shelf
Data centers typically contain many thin shelves, 1U servers mounted horizontally. Air is drawn on the front of the rack and runs out in the back. Because data centers typically contain a large number of computers and other power-relieving devices, they risk overheating equipment; Extensive HVAC systems are used to prevent this. Often elevated floors are used so that the area below the floor can be used as a large conference room for air cooling and electrical wiring.
Another way to accommodate a large number of systems in a small space is to use a blade chassis, which is oriented vertically rather than horizontally, to facilitate convection. Heated air by heat components tends to rise, creating natural airflow along the board (stack effect), cooling it down. Some manufacturers take advantage of this effect.
Laptop
Laptops present a difficult mechanical airflow design, power dissipation, and cooling challenges. Specific restrictions on laptops include: the overall device should be as light as possible; form factor should be built around standard keyboard layout; the user is very close, so the noise should be kept to a minimum, and the exterior temperature of the case must remain low enough to be used in rotation. Cooling generally uses forced air cooling but heat pipes and the use of metal chassis or casing as passive heat sinks are also common. Solutions to reduce heat include using lower ARM power consumption or Intel Atom processors.
Mobile devices
Mobile devices typically do not have discrete cooling systems, because the mobile CPU and GPU chips are designed for maximum power efficiency due to device battery constraints. Some of the higher performance devices may include a heat spreader that helps move heat to the outer casing of the phone or tablet.
See also
- CPU power dissipation
- Thermal design power
- Thermal management of devices and electronic systems
References
External links
- CPU Cooler Rule from the Thumb
- Submersion Cooling Patent Application
- DIY Submersion Cooling Gametrailers.com Forum - Video [1]. [2], [3].
- The Harsh Data and Environment Center uses commercially available Submersion Cooling at LiquidCool Solutions.
- Soaked Oil PC rig- Arctic Silver 5 test/mineral oil conductivity test. Text article/Video Driver YouTube- Oil-Oil Cooled PC Test Rig
- "Microsoft's new way to cool its data center: Throw them overboard".
Source of the article : Wikipedia
0 Comments