自旋轨道力矩磁性隧道结的物理、材料和器件应用

Such artificial intelligence as generative neuron networks are booming, which allows spintronics as a spring of randomness to have immense chances to apply for. Based on the studies on the field-free spin-orbit torque (SOT) effect [1-3] and successful development of high-performance SOT-magnetic tunnel junctions (MTJ) [4-5], we have been investigating applicability of SOT-MTJ as stochastic samplers in stochastic neuron networks [6, 7] and stochastic computing such as the restricted Boltzmann machines (RBM) prevalent in unsupervised learning and combinational problem solvers. Their works show SOT-MTJs well match the needs of RBM nodes, enabling the SOT-MTJ-sampled RBM to achieve handwritten and spoken digits recognition, generation and crossmodal learnings [8]. Their works clearly demonstrate spintronic devices ready for developing hardware tailored for stochastic networks and also open a promising outlet for spintronics, especially, SOT devices.

中国科学院物理研究所副研究员， 2007 年毕业于浙江大学； 2012 年获清华大学博士学位； 2012-2016 年在中科院物理所从事博士后研究工作； 2016 年在中科物理所任助研、副研至今。主要从事自旋电子学领域的研究工作，研究兴趣主要集中在自旋轨道电子学、磁子电子学和自旋热电子学等凝聚态磁学领域和磁随机存储器、自旋逻辑等自旋电子器件研发领域，近5年在Nature Electronics、Nature Communications、 Science Advances、 Phys. Rev. Lett.、 Advanced Materials、Nano Letters、Advanced Science等期刊上发表相关 SCI 论文 80 余篇，申请和获得中国发明专利 6 项，授权美国发明专利 1 项，获得中国科学院青年创新促进会项目资助。现已承担自然科学基金项目4项。