Small magnetic structures are being developed for magnetic sensors (for example, exploiting the giant magnetoresistance effect, or tunneling magnetoresistance) for nonvolatile memories exploiting either the Hall effect or magnetotransport effects and for storage on a nano scale. They present challenges in design for controlling the value and reproducibility of the switching field. Temperature effects as well as fluctuation of the grain structure on a mesoscopic scale become important. At the moment, magnetic nonvolatile memories are being developed by IBM-Siemans (Infinion) and the Motorola Companies. The uniformity of the switching field is a major stumbling block in the commercilaiztaion of high density nonvolatile magnetic memories. At the moment the yield of addressable sites is very small. Nearly all current micromagnetics calculations are carried out at zero temperature and do not include the temperature effects. In addition, very few systematic investigation were carried out. We propose to study and develop a way to predict the device characteristics and enable their optimum design with techniques that we have developed over the last several years. These include (a) calculations from an analytic point of view in combination with a finite temperature Monte Carlo code that we have written and optimized; (b) experimental studies of the switching field characteristics of our in-house e-beam patterned submicron structures.
