Thomas Windbacher was born in Mödling, Austria, in 1979. He studied physics at the Technische Universität Wien, where he received the degree of Diplomingenieur in October 2006. He joined the Institute for Microelectronics in October 2006 and finished his doctoral degree on engineering gate stacks for field-effect transistors in 2010. From 2010 until the beginning of 2012 he worked as a patent attorney candidate in Leoben. In March 2012 he rejoined the Institute for Microelectronics, where he currently works on the modeling and simulation of magnetic device structures.
Exploiting the Spin Hall Effect for the Improvement of a Non-Volatile Magnetic Flip-Flop
The endless demand for cheap and powerful electronics has propelled the scaling efforts of the semiconductor industry for many decades. Despite the ingenuity of researchers and engineers, who have made continuous scaling possible, the dimensions of CMOS devices have reached a point where both physical limits and a steep increase in factory costs for each subsequent technology node will soon bring scaling to an end. Therefore, in order to ensure future progress, it is necessary to investigate alternative materials, devices, and computational principles. At the same time, it is of the utmost importance that power consumption, interconnect delay, durability, fast operation, and long retention times are ensured for each new generation of technology.
The introduction of non-volatile elements, such as magnetic tunnel junctions and spin valves, into state-of-the-art computing systems, is a promising way to circumvent power dissipation and interconnect delay bottlenecks. This led us to propose a non-volatile magnetic flip-flop device, which offers high integration density, as well as CMOS compatibility. The device acts not only as an auxiliary memory element but also carries out the actual computation in the magnetic domain without relying on additional CMOS components.
The switching energy required, however, is still relatively high in comparison to CMOS, which necessitates a high current density for the flip-flop manipulation. We therefore proposed a modified device structure with a different device operation principle to benefit from the spin Hall effect in order to reduce the required switching energy without degrading other important parameters, such as switching speed or device reliability (cf. Figs. 1a and 1b). Our results show that the use of the spin Hall effect is rewarded by a simultaneous reduction in switching time (×5−×2) and switching energy (×5−×1.6) in comparison to a spin-transfer torque only device.
Fig. 1: In a first step, through both horizontal bars, a charge current is applied and through the Spin Hall effect generated spin accumulations, a torque is created that pulls the magnetization of the free layer towards in-plane. Then, two polarity encoded subsequent pulses, through the polarizer stacks, generate a spin transfer torque that either holds the initial magnetization orientation or switches the free layers' magnetization orientation.