skip to main content
DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information

This content will become publicly available on January 1, 2020

Title: Training a Neural Network on Analog TaO x ReRAM Devices Irradiated With Heavy Ions: Effects on Classification Accuracy Demonstrated With CrossSim

Abstract

The image classification accuracy of a TaO x ReRAM-based neuromorphic computing accelerator is evaluated after intentionally inducing a displacement damage up to a fluence of 10 14 2.5-MeV Si ions/cm 2 on the analog devices that are used to store weights. Results are consistent with a radiation-induced oxygen vacancy production mechanism. When the device is in the high-resistance state during heavy ion radiation, the device resistance, linearity, and accuracy after training are only affected by high fluence levels. Here, the findings in this paper are in accordance with the results of previous studies on TaO x -based digital resistive random access memory. When the device is in the low-resistance state during irradiation, no resistance change was detected, but devices with a 4- kΩ inline resistor did show a reduction in accuracy after training at 10 14 2.5-MeV Si ions/cm 2. This indicates that changes in resistance can only be somewhat correlated with changes to devices’ analog properties. This paper demonstrates that TaO x devices are radiation tolerant not only for high radiation environment digital memory applications but also when operated in an analog mode suitable for neuromorphic computation and training on new data sets.

Authors:
ORCiD logo [1];  [1]; ORCiD logo [1];  [1]; ORCiD logo [1];  [1];  [1];  [1]; ORCiD logo [2]; ORCiD logo [2]; ORCiD logo [1]
  1. Sandia National Lab. (SNL-NM), Albuquerque, NM (United States)
  2. Arizona State Univ., Tempe, AZ (United States). School of Electrical, Computer and Energy Engineering
Publication Date:
Research Org.:
Sandia National Lab. (SNL-NM), Albuquerque, NM (United States); Sandia National Laboratories, Livermore, CA (United States)
Sponsoring Org.:
USDOE National Nuclear Security Administration (NNSA)
OSTI Identifier:
1492357
Report Number(s):
SAND-2018-14169J
Journal ID: ISSN 0018-9499; 671088
Grant/Contract Number:  
AC04-94AL85000; NA0003525
Resource Type:
Accepted Manuscript
Journal Name:
IEEE Transactions on Nuclear Science
Additional Journal Information:
Journal Volume: 66; Journal Issue: 1; Journal ID: ISSN 0018-9499
Publisher:
IEEE
Country of Publication:
United States
Language:
English
Subject:
42 ENGINEERING; tantalum oxide; radiation effects; neuromorphic accelerator; Analog; deep learning

Citation Formats

Jacobs-Gedrim, Robin B, Hughart, David Russell, Agarwal, Sapan, Vizkelethy, Gyorgy, Bielejec, Edward S., Vaandrager, Bastiaan Leendert, Swanson, Scot E., Knisely, Katherine, Taggart, Jennifer L., Barnaby, Hugh L., and Marinella, Matthew J. Training a Neural Network on Analog TaOx ReRAM Devices Irradiated With Heavy Ions: Effects on Classification Accuracy Demonstrated With CrossSim. United States: N. p., 2019. Web. doi:10.1109/TNS.2018.2886229.
Jacobs-Gedrim, Robin B, Hughart, David Russell, Agarwal, Sapan, Vizkelethy, Gyorgy, Bielejec, Edward S., Vaandrager, Bastiaan Leendert, Swanson, Scot E., Knisely, Katherine, Taggart, Jennifer L., Barnaby, Hugh L., & Marinella, Matthew J. Training a Neural Network on Analog TaOx ReRAM Devices Irradiated With Heavy Ions: Effects on Classification Accuracy Demonstrated With CrossSim. United States. doi:10.1109/TNS.2018.2886229.
Jacobs-Gedrim, Robin B, Hughart, David Russell, Agarwal, Sapan, Vizkelethy, Gyorgy, Bielejec, Edward S., Vaandrager, Bastiaan Leendert, Swanson, Scot E., Knisely, Katherine, Taggart, Jennifer L., Barnaby, Hugh L., and Marinella, Matthew J. Tue . "Training a Neural Network on Analog TaOx ReRAM Devices Irradiated With Heavy Ions: Effects on Classification Accuracy Demonstrated With CrossSim". United States. doi:10.1109/TNS.2018.2886229.
@article{osti_1492357,
title = {Training a Neural Network on Analog TaOx ReRAM Devices Irradiated With Heavy Ions: Effects on Classification Accuracy Demonstrated With CrossSim},
author = {Jacobs-Gedrim, Robin B and Hughart, David Russell and Agarwal, Sapan and Vizkelethy, Gyorgy and Bielejec, Edward S. and Vaandrager, Bastiaan Leendert and Swanson, Scot E. and Knisely, Katherine and Taggart, Jennifer L. and Barnaby, Hugh L. and Marinella, Matthew J.},
abstractNote = {The image classification accuracy of a TaOx ReRAM-based neuromorphic computing accelerator is evaluated after intentionally inducing a displacement damage up to a fluence of 1014 2.5-MeV Si ions/cm2 on the analog devices that are used to store weights. Results are consistent with a radiation-induced oxygen vacancy production mechanism. When the device is in the high-resistance state during heavy ion radiation, the device resistance, linearity, and accuracy after training are only affected by high fluence levels. Here, the findings in this paper are in accordance with the results of previous studies on TaOx -based digital resistive random access memory. When the device is in the low-resistance state during irradiation, no resistance change was detected, but devices with a 4- kΩ inline resistor did show a reduction in accuracy after training at 1014 2.5-MeV Si ions/cm2. This indicates that changes in resistance can only be somewhat correlated with changes to devices’ analog properties. This paper demonstrates that TaOx devices are radiation tolerant not only for high radiation environment digital memory applications but also when operated in an analog mode suitable for neuromorphic computation and training on new data sets.},
doi = {10.1109/TNS.2018.2886229},
journal = {IEEE Transactions on Nuclear Science},
number = 1,
volume = 66,
place = {United States},
year = {2019},
month = {1}
}

Journal Article:
Free Publicly Available Full Text
This content will become publicly available on January 1, 2020
Publisher's Version of Record

Save / Share: