Shunsuke Fukami 교수

Spin-orbit torque (SOT)-induced magnetization switching shows promise for realizing ultrafast and reliable spintronics devices. Bipolar switching of perpendicular magnetization via SOT is achieved under an in-plane magnetic field collinear with an applied current. Typical structures studied so far comprise a nonmagnet/ferromagnet (NM/FM) bilayer, where the spin Hall effect in the NM is responsible for the switching. Here we show that an antiferromagnet/ferromagnet (AFM/FM) bilayer system also ex

Abstract We demonstrate associative memory operations reminiscent of the brain using nonvolatile spintronics devices. Antiferromagnet–ferromagnet bilayer-based Hall devices, which show analogue-like spin–orbit torque switching under zero magnetic fields and behave as artificial synapses, are used. An artificial neural network is used to associate memorized patterns from their noisy versions. We develop a network consisting of a field-programmable gate array and 36 spin–orbit torque devices. An e

Efficient information processing in the human brain is achieved by dynamics of neurons and synapses, motivating effective implementation of artificial spiking neural networks. Here, the dynamics of spin-orbit torque switching in antiferromagnet/ferromagnet heterostructures is studied to show the capability of the material system to form artificial neurons and synapses for asynchronous spiking neural networks. The magnetization switching, driven by a single current pulse or trains of pulses, is e

Current driven domain wall motion in nanostrips with perpendicular magnetic anisotropy was analyzed by using micromagnetic simulation. The threshold current density of perpendicular anisotropy strips in adiabatic approximation was much smaller than that of in-plane anisotropy strips, and it reduced with thickness reduction. The differences originate from the differences in domain wall width and hard-axis anisotropy. Also, the threshold current density of perpendicular anisotropy strips required

Current-induced domain wall motion in perpendicularly magnetized CoFeB nanowires with a stack structure of Ta(1.0 nm)/CoFeB(1.2 nm)/MgO(2.0 nm)/Ta(1.0 nm) was investigated. Domain wall motion driven by adiabatic spin-transfer torque was observed at a current of about 74 μA, corresponding to a current density of 6.2×107 A/cm2. The obtained results were compared with those of a micromagnetic simulation and the spin polarization of the CoFeB was estimated to be 0.72.

Magnetic heterostructures consisting of high-resistivity (238 ± 5 µΩ cm)-W/CoFeB/MgO are prepared by sputtering and their spin-orbit torques are evaluated as a function of W thickness through an extended harmonic measurement. W thickness dependence of the spin-orbit torque with the Slonczewski-like symmetry is well described by the drift-diffusion model with an efficiency parameter, the so-called effective spin Hall angle, of −0.62 ± 0.03. In contrast, the field-like spin-orbit torque is one ord

While digital integrated circuits with von Neumann architectures, having exponentially evolved for half a century, are an indispensable building block of today's information society, recently growing demand on executing more complex tasks like the human brain has allowed a revisit to the architecture of information processing. Brain-inspired hardware using artificial neural networks is expected to offer a complementary approach to deal with complex problems. Since the neuron and synapse are key

First Page

While artificial intelligence, capable of readily addressing cognitive tasks, has transformed technologies and daily lives, there remains a huge gap with biological systems in terms of performance per energy unit. Neuromorphic computing, in which hardware with alternative architectures, circuits, devices, and/or materials is explored, is expected to reduce the gap. Antiferromagnetic spintronics could offer a promising platform for this scheme. Active functionalities of antiferromagnetic systems

We have developed a new magnetic random access memory with current-induced domain wall (DW) motion (DW-motion MRAM). We confirmed its potential of 0.1-mA and 2-ns writing with sufficient thermal stability. The obtained properties indicate that this MRAM can replace conventional high-speed embedded memories.