Magnon Computing Algorithmics

PI: John Gregg

Department: Physics

Contemporary computing hardware faces a serious problem: it overheats. Little further performance improvement is possible until this problem is solved. Fortunately, a new technology is imminent that solves the three underlying issues: by offering computer clock speeds up to terahertz (representing trillions of pulses per second), heat dissipation reduced by a factor of 1000, and atomic scale components. This technology - Magnon Computing - not only holds the key to replicating conventional computational architecture with multiple performance improvements, but also has the potential to enable a computing paradigm shift. Magnonics offers the possibility of hardware that moves beyond conventional digital logic and hence removes the traditional divide between analogue and digital computing, thereby allowing us to combine the power and speed of the former with the versatility of the latter. In turn, such hybrid computer architectures will permit implementation of new algorithms capable of routinely performing tasks that are currently computationally challenging. Our very focused aim is to apply this breakthrough to enhance the power and speed of certain computationally expensive calculations.