|
Posted: May 7th, 2012 Honeycombs of magnets could lead to new type of computer processing ( Nanowerk News ) Scientists have taken an important step forward in developing anew material using nano-sized magnets that could ultimately lead tonew types of electronic devices, with greater processing capacitythan is currently feasible, in a study published in the journal Science ( "Emerging Chirality in Artificial Spin Ice" ). Many modern data storage devices, like hard disk drives, rely onthe ability to manipulate the properties of tiny individualmagnetic sections, but their overall design is limited by the waythese magnetic 'domains' interact when they are close together. Now, researchers from Imperial College London have demonstratedthat a honeycomb pattern of nano-sized magnets, in a material knownas spin ice, introduces competition between neighbouring magnets,and reduces the problems caused by these interactions bytwo-thirds. They have shown that large arrays of these nano-magnetscan be used to store computable information. The arrays can then beread by measuring their electrical resistance. Research author Dr Will Branford and his team have been investigating how to manipulate themagnetic state of nano-structured spin ices using a magnetic fieldand how to read their state by measuring their electricalresistance. They found that at low temperatures (below minus 223C)the magnetic bits act in a collective manner and arrange themselvesinto patterns. This changes their resistance to an electricalcurrent so that if one is passed through the material, thisproduces a characteristic measurement that the scientists canidentify. The scientists have so far been able to 'read' and 'write' patternsat room temperature. However, at the moment the collectivebehaviour is only seen at temperatures below minus 223C. A keychallenge now is to develop a way to utilise these patterns toperform calculations, and to do so at room temperature. Honeycomb shaped nano-magnet mesh. Current technology uses one magnetic domain to store a single bitof information. The new finding suggests that a cluster of manydomains could be used to solve a complex computational problem in asingle calculation. Computation of this type is known as a neuralnetwork, and is more similar to how our brains work than to the wayin which traditional computers process information. Dr Branford, who is an EPSRC Career Acceleration Fellow in theDepartment of Physics at Imperial College London, said:"Electronics manufacturers are trying all the time to squeeze moredata into the same devices, or the same data into a tinier spacefor handheld devices like smart phones and mobile computers.However, the innate interaction between magnets has so far limitedwhat they can do. In some new types of memory, manufacturers try toavoid the limitations of magnetism by avoiding using magnetsaltogether, using things like ferroelectric (flash) memory,memristors or antiferromagnets instead. However, these solutionsare slow, expensive or hard to read out. Our philosophy is toharness the magnetic interactions, making them work in our favour." Although today's research represents a key step forward, theresearchers say there are many hurdles to overcome beforescientists will be able to create prototype devices based on thistechnique such as developing an algorithm to control thecomputation. The nature of this algorithm will determine whetherthe room temperature state can be used or if the low temperaturecollective behaviour is required. However, they are optimistic thatif these challenges can be tackled successfully, new technologyusing magnetic honeycombs might be available in ten to fifteenyears. In experiments, Dr Branford applied an electrical current across acontinuous honeycomb mesh, made from cobalt magnetic bars each 1micrometer long and 100 nanometres wide, and covering an area 100square micrometers (as pictured). A single unit of the honeycombmesh is like three bar magnets meeting in the centre of a triangle.There is no way to arrange them without having either two northpoles or two south poles touching and repelling each other, this iscalled a 'frustrated' magnetic system. In a single triangular unitthere are six ways to arrange the magnets such that they haveexactly the same level of frustration, and as you increase thenumber of triangular units in the honeycomb, the number of possiblearrangements of magnets increases exponentially, increasing thecomplexity of possible patterns. Previous studies have shown that external magnetic fields can causethe magnetic domain of each bar to change state. This in turnaffects the interaction between that bar and its two neighbouringbars in the honeycomb, and it is this pattern of magnetic statesthat Dr Branford says could be computer data. Dr Branford said: "The strong interaction between neighbouringmagnets allows us to subtly affect how the patterns form across thehoneycomb. This is something we can take advantage of to computecomplex problems because many different outcomes are possible, andwe can differentiate between them electronically. Our next bigchallen ge i s to make an array of nano-magnets that can be'programmed' without using external magnetic fields.". We are high quality suppliers, our products such as White Sewing Thread , China Leather Sewing Thread for oversee buyer. To know more, please visits Leather Sewing Thread.
Related Articles -
White Sewing Thread, China Leather Sewing Thread,
|
