Improved reliability of magnetic field programmable gate arrays through the use of memristive tunnel junctions

Abstract

Since the recent successful implementation of the long-hypothesized memristor, its use in neuronal computing and in the reproduction of biological neural networks has gained increasing attention. In addition to the development of these new applications, the growing number of devices with memristive properties is promising to improve already established technologies. Herein, we use the recently reported memristance in magnesium-oxide-based magnetic tunnel junctions to improve the error tolerance in magnetic random access memory and magnetic field programmable logic arrays.