Design space exploration of Ferroelectric FET based Processing-in-Memory DNN Accelerator

In this letter, we quantify the impact of device limitations on the classification accuracy of an artificial neural network, where the synaptic weights are implemented in a Ferroelectric FET (FeFET) based in-memory processing architecture. We explore a design-space consisting of the resolution of the analog-to-digital converter, number of bits per FeFET cell, and the neural network depth. We show how the system architecture, training models and overparametrization can address some of the device limitations.

[1]  Gregory Cohen,et al.  EMNIST: an extension of MNIST to handwritten letters , 2017, CVPR 2017.

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

[3]  Yiran Chen,et al.  Memristor crossbar based hardware realization of BSB recall function , 2012, The 2012 International Joint Conference on Neural Networks (IJCNN).

[4]  J. Orbach.  Principles of Neurodynamics. Perceptrons and the Theory of Brain Mechanisms. , 1962 .

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

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

[7]  Klaus-Robert Müller,et al.  Neural Networks: Tricks of the Trade, this book is an outgrowth of a 1996 NIPS workshop , 1998, NIPS 1998.

[8]  Andrew L. Maas.  Rectifier Nonlinearities Improve Neural Network Acoustic Models , 2013 .

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

[10]  Thomas Mikolajick,et al.  About the deformation of ferroelectric hystereses , 2014 .

[11]  Shimeng Yu,et al.  Scaling-up resistive synaptic arrays for neuro-inspired architecture: Challenges and prospect , 2015, 2015 IEEE International Electron Devices Meeting (IEDM).

[12]  Yu Wang,et al.  Training itself: Mixed-signal training acceleration for memristor-based neural network , 2014, 2014 19th Asia and South Pacific Design Automation Conference (ASP-DAC).

[13]  Farnood Merrikh-Bayat,et al.  Training and operation of an integrated neuromorphic network based on metal-oxide memristors , 2014, Nature.

[14]  Wei Yang Lu,et al.  Nanoscale memristor device as synapse in neuromorphic systems. , 2010, Nano letters.

[15]  Shimeng Yu,et al.  Fully parallel write/read in resistive synaptic array for accelerating on-chip learning , 2015, Nanotechnology.

[16]  Chris Yakopcic,et al.  On-chip training of memristor crossbar based multi-layer neural networks , 2017, Microelectron. J..