A leading candidate for continued progress in the fields of information processing and communication once Moore\'s law is saturated, is the development a full spintronics technology based on using the spin instead of the charge of the electron (or hole) as the information carrier. Current spintronics research includes not only fundamental investigations of the quantum mechanical nature of spin in the hope that this will eventually lead to quantum data processing capabilities, but also the search for novel spin based devices where spin is treated at a more semi-classical level. In this presentation I will focus on the later aspect, and describe some of our results which we hope will lead to devices offering new functionalities and improved performance over conventional electronics.
I will begin by discussing issues which were at the forefront in the early development of spintronics, with a special focus on the problem on spin injection into semiconductors, and how that problem has been theoretically understood and experimentally overcome. I will then focus on our recent work on GaMnAs devices, where we have discovered interesting new spintronics functionalities, based on a novel magnetoresistance which we have dubbed tunneling anisotropic magnetoresistance (TAMR), and which can lead to changes of device resistance of several orders of magnitude as the direction of the magnetization vector is modified.