But a fan also consumes power — for a laptop, that is an extra drain on the battery. And fans alone are not always sufficient to cool the computer arrays used in data centers, many of which rely on heat exchangers that use liquid to cool the air flowing over the hot chips.
Image: Nature magazine & Ellsworth, M. J. et al. ITHERM 266–274 (2008)
Still larger machines demand more drastic measures. As Bruno Michel, manager of the advanced thermal packaging group at IBM in Rüschlikon, Switzerland, explains: “An advanced supercomputer would need a few cubic kilometers of air for cooling per day.” That simply is not practical, so computer engineers must resort to liquid cooling instead.
Water-cooled computers were commercially available as early as 1964, and several generations of mainframe computers built in the 1980s and 1990s were cooled by water. Today, non-aqueous, non-reactive liquid coolants such as fluorocarbons are sometimes used, often coming into direct contact with the chips. These substances generally cool by boiling — they absorb heat and the vapor carries it away. Other systems involve liquid sprays or refrigeration of the circuitry.
SuperMUC, an IBM-built supercomputer housed at the Leibniz center, became operational in 2012. The 3-petaflop machine is one of the world's most powerful supercomputers. It has a water-based cooling system, but the water is warm — around 45 °C. The water is pumped through microchannels carved into a customized copper heat sink above the central processing unit, which concentrates cooling in the parts of the system where it is most needed. The use of warm water may seem odd, but it consumes less energy than other cooling methods, because the hot water that emerges from the system requires less chilling before it is reintroduced. And the use of hot-water outflow for heating nearby office buildings results in further energy savings.
Michel and his colleagues at IBM believe that flowing water could be used not just to extract heat, but also to provide power for the circuitry in the first place, by carrying dissolved ions that engage in electrochemical reactions at energy-harvesting electrodes. In effect, the coolant doubles as an electrolyte 'fuel'. The idea is not entirely new, says Yogendra Joshi, a mechanical engineer at the Georgia Institute of Technology in Atlanta. “It has been used for many years in thermal management of aircraft electronics”, which are cooled by jet fuel, he says.
Delivering electrical power with an electrolyte flow is already a burgeoning technology. In a type of fuel cell known as a redox flow battery, for example, two electrolyte solutions are pumped into an electrochemical cell, where they are kept separate by a membrane that ions can flow through. Electrons travel between ions in the solutions in a process known as a reduction–oxidation (redox) reaction — but they are forced to do so through an external circuit, generating energy that can be tapped to provide electrical power.
Salty logic
Redox-flow cells can be miniaturized using microfluidic technology, in which the fluid flows are confined to microscopic channels etched into a substrate such as silicon. At such small scales, the liquids can flow past each other without mixing, so there is no need for a membrane to separate them. With this simplification, the devices are easier and cheaper to make, and they are compatible with silicon-chip technology.
Michel and his colleagues have begun to develop microfluidic cells for powering microprocessors, using a redox process based on vanadium ions. The electrolyte is pumped along microchannels that are 100–200 micrometers wide and similar to those used to carry coolant flows around some chips. Power is harvested at electrodes spaced along the channel, then distributed to individual devices by conventional metal wiring. The researchers unveiled their preliminary results in August, at a meeting of the International Society of Electrochemistry in Prague.
But they remain some way from actually powering circuits this way. At present, the power density of microfluidic redox-flow cells is less than 1 watt per square centimeter at 1 volt — two or three orders of magnitude too low to drive today's microprocessors. However, Michel believes that future processors will have significantly lower power requirements. And, he says, delivering power with microfluidic electrochemical cells should at least halve the power losses that occur with conventional metal wiring, which squanders around 50% of the electrical energy it carries as resistive heating.
See what we're tweeting about
5 Comments
Add CommentIf a chip is 2-D, but with devices on both sides, then devices on both sides can be cooled, for example, by emmersion in a flowing liquid. If most of the heat is in the wires, then put the wires on the outside where they can be cooled. While a 3-D architecture lets devices be closer to each other, you have the issue of how to cool stuff that isn't on the surface. One possible way of fixing that is to introduce holes for coolant to pass through the body of the device. How small can they be made? Can all devices be close enough to coolant? Does the liquid have to be forced through small tubes under high pressure? Perhaps a study of capilaries can help.
Reply | Report Abuse | Link to thisI dont believe in engineers !
Reply | Report Abuse | Link to thisEniac was NOT the first "modern electronic computer" and that was "built at the end of WW2".
Reply | Report Abuse | Link to thisThe first was built at Bletchley Park in Britain DURING WW2, and was instrumental in the code-breaking work that was carried out there on coded German radio transmissions.
A solution for cooling massive array's of computers circuits looms as the single largest obstacle to computing's future may have been solved using multiple low orbit processing centers cooled by space and powered by solar radiation? Computing using Low Orbit remote data processing centers that utilize encrypted microwave transmission to user centers for distributions to subscribers is the future?
Reply | Report Abuse | Link to thisGoogle could save $100's of Billions over Ten years of operations in energy costs it will spend to run current Data center and cooling costs. Space data centers would allow for the immediate development of future computing, exaflop machine.
Ron Nussbeck
Great article - thanks. I was wondering if to the best of your knowledge anyone is actively working on a "reversible computing" means as suggested by Ray Kurzweil? That method is supposed to generate zero net thermal loss. I can sorta-kinda envisage the logical operations involved - and wondered if perhaps the new memristor semiconductor flavor or the much in-the-news electron spin storage and other pending quantum computing widgets would get us any closer to that dream?
Reply | Report Abuse | Link to thisTo clarify the reference - quoting from his "The Singularity is Near"... Quote: "Rolf Landauer showed in 1961 that reversible logical operations such as NOT (turning a bit into its opposite) could be performed without putting energy in or taking heat out, but that irreversible logical operations such as AND (generating bit C, which is a 1 if and only if both inputs A and Bare 1) do require energy. In 1973 Charles Bennett showed that any computation could be performed using only reversible logical operations. A decade later, Ed Fredkin and Tommaso Toffoli presented a comprehensive review of the idea of reversible computing. The fundamental concept is that if you keep all the intermediate results and then run the algorithm backward when you've finished your calculation, you end up where you started, have used no energy, and generated no heat. Along the way, however, you've calculated the result of the algorithm." End quote.
He then goes on to inspiringly ask "How Smart Is a Rock?". Wondering if there any fundamental efforts towards changing the calculation process, not just the wiring and location? (This middle aged engineer needs a zetaflop before he turns 100!).