Progress in spin-based intelligent devices and physical neural networks

<sec><p indent="0mm">In the era of information explosion, addressing computational bottlenecks and escalating energy demands has become crucial. Conventional computer systems face performance limitations, prompting the search for novel computational paradigms. Spin-based electronic devices, renowned for their non-volatility, fast response and high energy efficiency, offer promising solutions. This review outlines recent progress in spin-based intelligent devices and physical neural networks, focusing on spintronic synapses, spin oscillator neurons, spin-based artificial neural networks (ANNs), and probabilistic neural networks. </sec><sec> Spin-orbit torque (SOT) devices are emerging as attractive candidates for artificial synapses due to their non-volatility and tunability. By exploiting methods like interlayer exchange coupling, exchange bias, and magnetic field modulation, researchers have achieved stable multi-state switching in SOT devices, mimicking synaptic weight changes. This enables functionalities like spike-timing-dependent plasticity (STDP), critical for learning and information storage. Devices utilizing topological magnetic structures like skyrmions show linear weight updates, enhancing performance. </sec><sec> Inspired by the nonlinear dynamics of biological neurons, spin-torque nano-oscillators (STNOs) and spin-Hall nano-oscillators (SHNOs) exhibit similar behavior. STNOs utilize spin-transfer torque, while SHNOs leverage spin Hall effect for microwave oscillations. Applications in recurrent neural networks (RNNs) for speech recognition and radiofrequency (RF) signal identification have demonstrated their potential in neuromorphic computing. </sec><sec> Building on the low power consumption and non-volatility of spin devices, spin-based ANNs have emerged as viable alternatives to CMOS-based networks. Arrays of spintronic devices facilitate in-memory computing, significantly reducing energy consumption. Studies using STT-MRAM and SOT devices have achieved binary and analog weight neural network computations, demonstrating their suitability for edge computing and IoT applications. </sec><sec> Recognizing the role of stochasticity in biological neural networks, spintronic devices with low energy barriers exhibit intrinsic randomness, enabling probabilistic computation. By manipulating the switching probability of magnetic tunnel junctions (MTJs), probabilistic bits mimic neurons capable of solving complex problems like Ising model optimization. High-energy barrier SOT devices retain non-volatility while allowing tunable stochasticity, advancing the field. </sec><sec> Recent advancements in spin-based devices have paved the way for the development of physical neural networks that mimic the functionalities of biological neural systems. By precisely controlling magnetic moments, researchers are designing circuits and chips that can dynamically adjust their connections, mirroring the cognitive and learning abilities of the brain. This groundbreaking technology promises profound impacts on neuromorphic computing, brain-inspired AI, and addressing the bottlenecks in current AI systems related to efficiency, energy consumption, and adaptability. </sec>