Machine Learning Approach for Device-Circuit Co-Optimization of Stochastic-Memristive-Device-Based Boltzmann Machine

A Boltzmann machine whose effective "temperature" can be dynamically "cooled" provides a stochastic neural network realization of simulated annealing, which is an important metaheuristic for solving combinatorial or global optimization problems with broad applications in machine intelligence and operations research. However, the hardware realization of the Boltzmann stochastic element with "cooling" capability has never been achieved within an individual semiconductor device. Here we demonstrate a new memristive device concept based on two-dimensional material heterostructures that enables this critical stochastic element in a Boltzmann machine. The dynamic cooling effect in simulated annealing can be emulated in this multi-terminal memristive device through electrostatic bias with sigmoidal thresholding distributions. We also show that a machine-learning-based method is efficient for device-circuit co-design of the Boltzmann machine based on the stochastic memristor devices in simulated annealing. The experimental demonstrations of the tunable stochastic memristors combined with the machine-learning-based device-circuit co-optimization approach for stochastic-memristor-based neural-network circuits chart a pathway for the efficient hardware realization of stochastic neural networks with applications in a broad range of electronics and computing disciplines.

[1]  Hyo Seon Park,et al.  Neurocomputing for Design Automation , 2018 .

[2]  Yiran Chen,et al.  PipeLayer: A Pipelined ReRAM-Based Accelerator for Deep Learning , 2017, 2017 IEEE International Symposium on High Performance Computer Architecture (HPCA).

[3]  Zbigniew J. Czech,et al.  Parallel simulated annealing for the vehicle routing problem with time windows , 2002, Proceedings 10th Euromicro Workshop on Parallel, Distributed and Network-based Processing.

[4]  Wang Yao,et al.  Optical signature of symmetry variations and spin-valley coupling in atomically thin tungsten dichalcogenides , 2012, Scientific Reports.

[5]  Madhavan Swaminathan,et al.  Preliminary application of machine-learning techniques for thermal-electrical parameter optimization in 3-D IC , 2016, 2016 IEEE International Symposium on Electromagnetic Compatibility (EMC).

[6]  Simon Haykin,et al.  Neural Networks and Learning Machines , 2010 .

[7]  Ehl Emile Aarts,et al.  Simulated annealing and Boltzmann machines , 2003 .

[8]  Jing Guo,et al.  Atomically Thin CBRAM Enabled by 2-D Materials: Scaling Behaviors and Performance Limits , 2018, IEEE Transactions on Electron Devices.

[9]  Andrzej Cichocki,et al.  Neural networks for optimization and signal processing , 1993 .

[10]  Donald R. Jones,et al.  A Taxonomy of Global Optimization Methods Based on Response Surfaces , 2001, J. Glob. Optim..

[11]  Jasper Snoek,et al.  Practical Bayesian Optimization of Machine Learning Algorithms , 2012, NIPS.

[12]  Yiran Chen,et al.  An overview of non-volatile memory technology and the implication for tools and architectures , 2009, 2009 Design, Automation & Test in Europe Conference & Exhibition.