Spintronic Neuromorphic Computing Systems

Spintronic neuromorphic computing systems exploit the electron’s spin alongside its charge to emulate the key functions of biological neural networks. By integrating memory and logic within magnetic devices, these architectures aim to overcome the conventional separation of storage and computation, known as the von Neumann bottleneck. Central elements include magnetic tunnel junctions, spin–orbit torque mechanisms, domain-wall conduits and memristive elements, all of which offer non-volatility, ultrafast switching and high endurance. Current-driven magnetic dynamics can implement synaptic weight updates and neuronal activation functions, enabling both spiking and analogue processing. Demonstrations to date have achieved multilevel weight tuning, deterministic and stochastic domain-wall motion, and in-memory arithmetic operations. Research continues to focus on material optimisation, device scalability and the seamless integration of spintronic synapses and neurons into prototype hardware networks for applications in pattern recognition, associative memory and beyond.

Research from Nature Portfolio

Recent studies have established multilevel spin-torque memristors as viable artificial synapses. By exploiting perpendicular magnetic anisotropy in magnetic tunnel junctions, researchers have achieved analogue resistance tuning through controlled domain-wall displacement. These devices exhibit a large number of stable intermediate states at low current densities, minimising energy consumption while maintaining high endurance. In parallel work, deterministic manipulation of ferromagnetic domain walls orthogonal to current flow has been demonstrated via spin-orbit torques alone. This orthogonal motion introduces a novel control modality for synaptic weight updates, enabling precise, reproducible adjustments in neuromorphic circuits without the need for external magnetic fields.

Research from all publishers

Integrated neuromorphic networks combining spintronic synapses and neurons have been realised, demonstrating electric-field-mediated weight updates and activation functions in a compact hardware prototype. A proof-of-concept pattern-classification task achieved over 93% accuracy using stripe-domain synapses for linear, symmetric weight modulation alongside spin-orbit-torque-driven neurons. Complementary advances employ voltage-controlled domain-wall motion in magnetic tunnel junctions to construct stochastic synapses and nonlinear neuronal units. By engineering intrinsic pinning sites and harnessing thermal fluctuations, this approach achieves potentiation and depression behaviours, culminating in a digit-recognition demonstration based on spike-time-dependent plasticity algorithms.

Spintronic Neuromorphic Computing Systems publication trend

The graph below shows the total number of articles in spintronic neuromorphic computing systems across all publications each year (not limited to Nature Index journals).

Technical terms

Spintronics: A field exploiting electron spin and charge to store and process information in magnetic materials.

Neuromorphic computing: Hardware architectures emulating neural systems to achieve parallel, low-power information processing.

Spin–orbit torque (SOT): A mechanism by which spin currents generate torques on magnetic moments, enabling current-driven magnetisation switching.

Magnetic tunnel junction (MTJ): A nanoscale device comprising two ferromagnetic layers separated by an insulating barrier, used for non-volatile memory and logic operations.

Domain wall: A boundary between regions of differing magnetisation orientation, movable by spin currents for information encoding.

Memristor: A two-terminal device whose resistance depends on the history of applied electrical stimuli, suitable for analogue weight storage.

References

1. A magnetic synapse: multilevel spin-torque memristor with perpendicular anisotropy. Scientific Reports (2016).
2. Deterministic Domain Wall Motion Orthogonal To Current Flow Due To Spin Orbit Torque. Scientific Reports (2015).
3. Integrated neuromorphic computing networks by artificial spin synapses and spin neurons. NPG Asia Materials (2021).
4. Magnetic Elements for Neuromorphic Computing. Molecules (2020).
5. Voltage-Controlled Domain Wall Motion-Based Neuron and Stochastic Magnetic Tunnel Junction Synapse for Neuromorphic Computing Applications. IEEE Journal on Exploratory Solid-State Computational Devices and Circuits (2021).
6. Spike time dependent plasticity (STDP) enabled learning in spiking neural networks using domain wall based synapses and neurons. AIP Advances (2019).
7. In‐Memory Mathematical Operations with Spin‐Orbit Torque Devices. Advanced Science (2022).