TY - JOUR
T1 - Evaluation of Radiation Effects in RRAM-Based Neuromorphic Computing System for Inference
AU - Ye, Zhilu
AU - Liu, Rui
AU - Taggart, Jennifer
AU - Barnaby, Hugh
AU - Yu, Shimeng
N1 - Funding Information:
Manuscript received October 26, 2018; revised November 27, 2018; accepted December 10, 2018. Date of publication December 14, 2018; date of current version January 17, 2019. This work was supported by DoD-DTRA under Grant HDTRA1-16-1-0012 and Grant HDTRA1-17-1-0038.
Publisher Copyright:
© 2018 IEEE.
PY - 2019/1
Y1 - 2019/1
N2 - Neuromorphic computing systems built with resistive random access memory (RRAM) are attractive solutions for implementing deep learning on-chip. In this paper, the single event and cumulative radiation damage susceptibility of a 1-transistor-1-resistor synaptic array architecture is investigated for the inference stage of a neuromorphic system. A physics-based SPICE RRAM model is developed to simulate the analog switching behavior of HfOx-based RRAM devices. SPICE simulations using the RRAM model are performed to capture the effects of transient heavy-ion strikes and experimental results obtained on GeSe-based RRAM devices are used to model the effects of cumulative radiation dose. Error patterns are fed into the software simulation of a multilayer perceptron, a representative artificial neural network for MNIST handwritten digit recognition. Radiation effects on the inference accuracy are analyzed. The results of the analysis indicate that the RRAM-based neuromorphic computing system is highly resistant to transient single-event effects due to the low cross section. The system is also shown to be tolerant to multi-Mrad levels of total ionizing and displacement damage dose.
AB - Neuromorphic computing systems built with resistive random access memory (RRAM) are attractive solutions for implementing deep learning on-chip. In this paper, the single event and cumulative radiation damage susceptibility of a 1-transistor-1-resistor synaptic array architecture is investigated for the inference stage of a neuromorphic system. A physics-based SPICE RRAM model is developed to simulate the analog switching behavior of HfOx-based RRAM devices. SPICE simulations using the RRAM model are performed to capture the effects of transient heavy-ion strikes and experimental results obtained on GeSe-based RRAM devices are used to model the effects of cumulative radiation dose. Error patterns are fed into the software simulation of a multilayer perceptron, a representative artificial neural network for MNIST handwritten digit recognition. Radiation effects on the inference accuracy are analyzed. The results of the analysis indicate that the RRAM-based neuromorphic computing system is highly resistant to transient single-event effects due to the low cross section. The system is also shown to be tolerant to multi-Mrad levels of total ionizing and displacement damage dose.
KW - Analog resistive random access memory (RRAM)
KW - multilayer perceptron (MLP)
KW - neuromorphic computing
KW - single-event upset (SEU)
KW - total ionizing dose (TID)
UR - http://www.scopus.com/inward/record.url?scp=85058899960&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85058899960&partnerID=8YFLogxK
U2 - 10.1109/TNS.2018.2886793
DO - 10.1109/TNS.2018.2886793
M3 - Article
AN - SCOPUS:85058899960
SN - 0018-9499
VL - 66
SP - 97
EP - 103
JO - IEEE Transactions on Nuclear Science
JF - IEEE Transactions on Nuclear Science
IS - 1
M1 - 8576612
ER -