Magnetic domain wall based synaptic and activation function generator for neuromorphic accelerators

Magnetic domain walls are information tokens in both logic and memory devices, and hold particular interest in applications such as neuromorphic accelerators that combine logic in memory. Here, we show that devices based on the electrical manipulation of magnetic domain walls are capable of implementing linear, as well as programmable nonlinear, functions. Unlike other approaches, domain-wall-based devices are ideal for application to both synaptic weight generators and thresholding in deep neural networks. Prototype micrometer-size devices operate with 8 ns current pulses and the energy consumption required for weight modulation is ≤ 16 pJ. Both speed and energy consumption compare favorably to other synaptic nonvolatile devices, with the expected energy dissipation for scaled 20 nm devices close to that of biological neurons.

[1]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[2]  S. Parkin,et al.  Chiral spin torque at magnetic domain walls. , 2013, Nature nanotechnology.

[3]  S. Parkin,et al.  Magnetic Domain-Wall Racetrack Memory , 2008, Science.

[4]  Saima Afroz Siddiqui,et al.  The Spatial Resolution Limit for an Individual Domain Wall in Magnetic Nanowires. , 2017, Nano letters.

[5]  Andrew Zisserman,et al.  Deep Convolutional Neural Networks for Efficient Pose Estimation in Gesture Videos , 2014, ACCV.

[6]  A. Fert,et al.  Field-free switching of a perpendicular magnetic tunnel junction through the interplay of spin–orbit and spin-transfer torques , 2018, Nature Electronics.

[7]  E. Linfield,et al.  Toward Chirality‐Encoded Domain Wall Logic , 2019, Advanced Functional Materials.

[8]  Geoffrey E. Hinton,et al.  Learning representations by back-propagating errors , 1986, Nature.

[9]  Luc Thomas,et al.  Discrete domain wall positioning due to pinning in current driven motion along nanowires. , 2011, Nano letters.

[10]  H.-S. Philip Wong,et al.  In-memory computing with resistive switching devices , 2018, Nature Electronics.

[11]  C. Ross,et al.  Current-Induced Domain Wall Motion in a Compensated Ferrimagnet. , 2018, Physical review letters.

[12]  D Petit,et al.  Magnetic Domain-Wall Logic , 2005, Science.

[13]  Shimeng Yu,et al.  Ferroelectric FET analog synapse for acceleration of deep neural network training , 2017, 2017 IEEE International Electron Devices Meeting (IEDM).

[14]  B. DeSalvo,et al.  CBRAM devices as binary synapses for low-power stochastic neuromorphic systems: Auditory (Cochlea) and visual (Retina) cognitive processing applications , 2012, 2012 International Electron Devices Meeting.

[15]  K. Xia,et al.  An all-metallic logic gate based on current-driven domain wall motion. , 2008, Nature nanotechnology.

[16]  Geoffrey E. Hinton,et al.  Deep Learning , 2015, Nature.

[17]  C. A. Ross,et al.  Logic circuit prototypes for three-terminal magnetic tunnel junctions with mobile domain walls , 2016, Nature Communications.

[18]  L. Pileggi,et al.  Experimental demonstration of four-terminal magnetic logic device with separate read- and write-paths , 2014, 2014 IEEE International Electron Devices Meeting.

[19]  C. Ross,et al.  Magnetic reversal and thermal stability of CoFeB perpendicular magnetic tunnel junction arrays patterned by block copolymer lithography , 2018, Nanotechnology.

[20]  Shimeng Yu,et al.  Stochastic learning in oxide binary synaptic device for neuromorphic computing , 2013, Front. Neurosci..

[21]  Anupam Chattopadhyay,et al.  Spintronic Device-Structure for Low-Energy XOR Logic using Domain Wall Motion , 2019, 2019 IEEE International Symposium on Circuits and Systems (ISCAS).

[22]  D. Edelstein,et al.  Key parameters affecting STT-MRAM switching efficiency and improved device performance of 400°C-compatible p-MTJs , 2017, 2017 IEEE International Electron Devices Meeting (IEDM).

[23]  Pritish Narayanan,et al.  Experimental Demonstration and Tolerancing of a Large-Scale Neural Network (165 000 Synapses) Using Phase-Change Memory as the Synaptic Weight Element , 2014, IEEE Transactions on Electron Devices.

[24]  Gökmen Tayfun,et al.  Acceleration of Deep Neural Network Training with Resistive Cross-Point Devices: Design Considerations , 2016, Front. Neurosci..

[25]  Rob R. de Ruyter van Steveninck,et al.  The metabolic cost of neural information , 1998, Nature Neuroscience.

[26]  Robert M. White,et al.  Two-terminal spin–orbit torque magnetoresistive random access memory , 2018, Nature Electronics.

[27]  H. Ohno,et al.  Current-induced domain wall motion in magnetic nanowires with various widths down to less than 20 nm , 2016 .

[28]  S. Yuasa,et al.  A magnetic synapse: multilevel spin-torque memristor with perpendicular anisotropy , 2016, Scientific Reports.

[29]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[30]  K. Roy,et al.  Spin-Based Neuron Model With Domain-Wall Magnets as Synapse , 2012, IEEE Transactions on Nanotechnology.

[31]  Caroline A. Ross,et al.  A logic-in-memory design with 3-terminal magnetic tunnel junction function evaluators for convolutional neural networks , 2017, 2017 IEEE/ACM International Symposium on Nanoscale Architectures (NANOARCH).

[32]  A. Pohm,et al.  Effect of Resistance-Area-Product and Thermal Environment on Writing of Magneto-Thermal MRAM , 2006, IEEE Transactions on Magnetics.

[33]  S. Fukami,et al.  Low-current perpendicular domain wall motion cell for scalable high-speed MRAM , 2006, 2009 Symposium on VLSI Technology.

[34]  H. Hwang,et al.  HfZrOx-Based Ferroelectric Synapse Device With 32 Levels of Conductance States for Neuromorphic Applications , 2017, IEEE Electron Device Letters.

[35]  Tara N. Sainath,et al.  Deep Neural Networks for Acoustic Modeling in Speech Recognition , 2012 .

[36]  Vivienne Sze,et al.  Efficient Processing of Deep Neural Networks: A Tutorial and Survey , 2017, Proceedings of the IEEE.

[37]  H. Ohno,et al.  Interface control of the magnetic chirality in CoFeB/MgO heterostructures with heavy-metal underlayers. , 2014, Nature communications.

[38]  Kaushik Roy,et al.  STT-SNN: A Spin-Transfer-Torque Based Soft-Limiting Non-Linear Neuron for Low-Power Artificial Neural Networks , 2014, IEEE Transactions on Nanotechnology.

[39]  H. Ohno,et al.  A perpendicular-anisotropy CoFeB-MgO magnetic tunnel junction. , 2010, Nature materials.