Spatial-Temporal Hybrid Neural Network With Computing-in-Memory Architecture
- 16 April 2021
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 68 (7) , 2850-2862
Abstract
Deep learning (DL) has gained unprecedented success in many real-world applications. However, DL poses difficulties for efficient hardware implementation due to the needs of a complex gradient-based learning algorithm and the required high memory bandwidth for synaptic weight storage, especially in today's data-intensive environment. Computing-in-memory (CIM) strategies have emerged as an alternative for realizing energy-efficient neuromorphic applications in silicon, reducing resources and energy required for neural computations. In this work, we exploit a CIM-based spatial-temporal hybrid neural network (STHNN) with a unique learning algorithm. To be specific, we integrate both multilayer perceptron and recurrent-based delay-dynamical system, making the network becomes linear separable while processing information in both spatial and temporal domains, better yet, reducing the memory bandwidth and hardware overhead through the CIM architecture. The prototype fabricated in 180 nm CMOS process is built of fully-analog components, yielding an average on-chip classification accuracy up to 86.9% on handprinted alphabet characters with a power consumption of 33 mW. Beyond that, through the handwritten digit database and the radio frequency fingerprinting dataset, software-based numerical evaluations offer 1.6 -to- 9.8 × and 1.9 -to- 4.4 × speedup, respectively, without significantly degrading its classification accuracy compared to the cutting-edge DL approaches.Keywords
Funding Information
- U.S. National Science Foundation (CCF-1750450, ECCS-1811497, CCF-1937487)
This publication has 43 references indexed in Scilit:
- NeuCube: A spiking neural network architecture for mapping, learning and understanding of spatio-temporal brain dataNeural Networks, 2014
- Parallel photonic information processing at gigabyte per second data rates using transient statesNature Communications, 2013
- A Large-Scale Model of the Functioning BrainScience, 2012
- Re-visiting the echo state propertyNeural Networks, 2012
- Information processing using a single dynamical node as complex systemNature Communications, 2011
- Real-Time Computing Without Stable States: A New Framework for Neural Computation Based on PerturbationsNeural Computation, 2002
- Gradient-based learning applied to document recognitionProceedings of the IEEE, 1998
- Long Short-Term MemoryNeural Computation, 1997
- High-dimensional chaotic behavior in systems with time-delayed feedbackPhysica D: Nonlinear Phenomena, 1987
- Oscillation and Chaos in Physiological Control SystemsScience, 1977