Spin memristors based on antiferromagnetic and altermagneticmaterials represent a frontier in neuromorphic and non-volatile memory technologies, offering compelling advantages over their ferromagnetic counterparts. Unlike ferromagnets, antiferromagnets possess zero net magnetization due to their antiparallel spin sublattice arrangement, rendering them largely immune to stray magnetic field perturbations and enabling devices with exceptional thermal stability and ultrafast switching dynamics in the terahertz regime. Altermagnets, a recently recognized magnetic phase distinct from both ferromagnets and conventional antiferromagnets, further enrich this landscape by combining vanishing net magnetization with momentum-dependent spin splitting — a property that gives rise to anomalous Hall and spin Hall responses without requiring spin-orbit coupling in the conventional sense. These characteristics make both material classes highly attractive for memristive applications, where resistance states must be reliably written, retained, and read across many cycles.
The operating principle of antiferromagnetic and altermagneticmemristors generally relies on the controlled manipulation of Néel order or spin-texture configurations through mechanisms such as spin-orbit torque, spin-transfer torque, or current-induced staggered fields, with resistance readout achieved via the tunneling magnetoresistance or anisotropic magnetoresistance effect. The multi-level resistance states accessible through partial switching of spin domains offer a natural analog to synaptic weight modulation, positioning these devices as strong candidates for hardware implementations of artificial neural networks. Altermagnets, with their symmetry-protected spin-split band structures in materials such as RuO₂, MnTe, and CrSb, add an additional degree of freedom by enabling electrical detection schemes that were previously unavailable in compensated magnets. Continued progress in epitaxial thin-film growth, interface engineering, and the theoretical understanding of Néel vector dynamics will be critical to translating these promising material properties into scalable, energy-efficient memristive devices.