The promise of spintronics for unconventional computing

Magnetic tunnel junctions (MTJs) with ferromagnetic electrodes possessing a perpendicular magnetic easy axis are of great interest as they have a potential for realizing next-generation high-density non-volatile memory and logic chips with high thermal stability and low critical current for current-induced magnetization switching. To attain perpendicular anisotropy, a number of material systems have been explored as electrodes, which include rare-earth/transition-metal alloys, L1(0)-ordered (Co, Fe)-Pt alloys and Co/(Pd, Pt) multilayers. However, none of them so far satisfy high thermal stability at reduced dimension, low-current current-induced magnetization switching and high tunnel magnetoresistance ratio all at the same time. Here, we use interfacial perpendicular anisotropy between the ferromagnetic electrodes and the tunnel barrier of the MTJ by employing the material combination of CoFeB-MgO, a system widely adopted to produce a giant tunnel magnetoresistance ratio in MTJs with in-plane anisotropy. This approach requires no material other than those used in conventional in-plane-anisotropy MTJs. The perpendicular MTJs consisting of Ta/CoFeB/MgO/CoFeB/Ta show a high tunnel magnetoresistance ratio, over 120%, high thermal stability at dimension as low as 40 nm diameter and a low switching current of 49 microA.

A key challenge for neural modeling is to explain how a continuous stream of multimodal input from a rapidly changing environment can be processed by stereotypical recurrent circuits of integrate-and-fire neurons in real time. We propose a new computational model for real-time computing on time-varying input that provides an alternative to paradigms based on Turing machines or attractor neural networks. It does not require a task-dependent construction of neural circuits. Instead, it is based on principles of high-dimensional dynamical systems in combination with statistical learning theory and can be implemented on generic evolved or found recurrent circuitry. It is shown that the inherent transient dynamics of the high-dimensional dynamical system formed by a sufficiently large and heterogeneous neural circuit may serve as universal analog fading memory. Readout neurons can learn to extract in real time from the current state of such recurrent neural circuit information about current and past inputs that may be needed for diverse tasks. Stable internal states are not required for giving a stable output, since transient internal states can be transformed by readout neurons into stable target outputs due to the high dimensionality of the dynamical system. Our approach is based on a rigorous computational model, the liquid state machine, that, unlike Turing machines, does not require sequential transitions between well-defined discrete internal states. It is supported, as the Turing machine is, by rigorous mathematical results that predict universal computational power under idealized conditions, but for the biologically more realistic scenario of real-time processing of time-varying inputs. Our approach provides new perspectives for the interpretation of neural coding, the design of experiments and data analysis in neurophysiology, and the solution of problems in robotics and neurotechnology.