Ionizing radiation effects in SONOS-based neuromorphic inference accelerators

Patrick Xiao, T.; Bennett, Christopher H.; Agarwal, Sapan; Hughart, David R.; Barnaby, Hugh J.; Puchner, Helmut; Prabhakar, Venkatraman; Alec Talin, A.; Marinella, Matthew J.

doi:10.1109/tns.2021.3058548

Ionizing radiation effects in SONOS-based neuromorphic inference accelerators

Journal Article · Wed Feb 10 23:00:00 EST 2021 · IEEE Transactions on Nuclear Science

DOI:https://doi.org/10.1109/tns.2021.3058548· OSTI ID:1770799

Patrick Xiao, T. ^[1]; Bennett, Christopher H. ^[1]; Agarwal, Sapan ^[1]; Hughart, David R. ^[1]; Barnaby, Hugh J. ^[2]; Puchner, Helmut ^[3]; Prabhakar, Venkatraman ^[3]; Alec Talin, A. ^[1]; Marinella, Matthew J. ^[1]

Sandia National Lab. (SNL-NM), Albuquerque, NM (United States)
Arizona State Univ., Tempe, AZ (United States)
Infineon Technologies, San Jose, CA (United States)

Here, we evaluate the sensitivity of neuromorphic inference accelerators based on Silicon-Oxide-Nitride-Oxide-Silicon (SONOS) charge trap memory arrays to total ionizing dose (TID) effects. Data retention statistics were collected for 16 Mbit of 40 nm SONOS digital memory exposed to ionizing radiation from a Co-60 source, showing good retention of the bits up to the maximum dose of 500 krad(Si). Using this data, we formulate a rate-equation-based model for the TID response of trapped charge carriers in the ONO stack, and predict the effect of TID on intermediate device states between ‘program’ and ‘erase’. This model is then used to simulate arrays of low-power, analog SONOS devices that store 8-bit neural network weights and support in situ matrix-vector multiplication. We evaluate the accuracy of the irradiated SONOS-based inference accelerator on two image recognition tasks – CIFAR-10 and the challenging ImageNet dataset – using state-of-the-art convolutional neural networks, such as ResNet-50. We find that across the datasets and neural networks evaluated, the accelerator tolerates a maximum TID between 10 krad(Si) and 100 krad(Si), with deeper networks being more susceptible to accuracy losses due to TID.

View Accepted Manuscript (DOE)

Research Organization:: Sandia National Laboratories (SNL-NM), Albuquerque, NM (United States)

Sponsoring Organization:: USDOE National Nuclear Security Administration (NNSA); Defense Threat Reduction Agency (DTRA)

Grant/Contract Number:: AC04-94AL85000; NA0003525

OSTI ID:: 1770799

Report Number(s):: SAND--2021-2579J; 694519

Journal Information:: IEEE Transactions on Nuclear Science, Journal Name: IEEE Transactions on Nuclear Science Journal Issue: 5 Vol. 68; ISSN 0018-9499

Publisher:: IEEECopyright Statement

Country of Publication:: United States

Language:: English

References (27)

Scaling of multidielectric nonvolatile SONOS memory structures French, Margaret L.; White, Marvin H. Solid-State Electronics, Vol. 37, Issue 12 https://doi.org/10.1016/0038-1101(94)90057-4	journal	December 1994
Accurate deep neural network inference using computational phase-change memory Joshi, Vinay; Le Gallo, Manuel; Haefeli, Simon Nature Communications, Vol. 11, Issue 1 https://doi.org/10.1038/s41467-020-16108-9	journal	May 2020
Electron trap density distribution of Si-rich silicon nitride extracted using the modified negative charge decay model of silicon-oxide-nitride-oxide-silicon structure at elevated temperatures Kim, Tae Hun; Park, Il Han; Lee, Jong Duk Applied Physics Letters, Vol. 89, Issue 6 https://doi.org/10.1063/1.2335619	journal	August 2006
Gap states in silicon nitride Robertson, John; Powell, Martin J. Applied Physics Letters, Vol. 44, Issue 4 https://doi.org/10.1063/1.94794	journal	February 1984
Neuromorphic computing using non-volatile memory Burr, Geoffrey W.; Shelby, Robert M.; Sebastian, Abu Advances in Physics: X, Vol. 2, Issue 1 https://doi.org/10.1080/23746149.2016.1259585	journal	October 2016
Total Ionizing Dose Effects of 55-nm Silicon-Oxide-Nitride-Oxide-Silicon Charge Trapping Memory in Pulse and DC Modes Li, Mei; Bi, Jin-Shun; Xu, Yan-Nan Chinese Physics Letters, Vol. 35, Issue 7 https://doi.org/10.1088/0256-307X/35/7/078502	journal	July 2018
Radiation response of floating gate EEPROM memory cells Snyder, E. S.; McWhorter, P. J.; Dellin, T. A. IEEE Transactions on Nuclear Science, Vol. 36, Issue 6 https://doi.org/10.1109/23.45415	journal	December 1989
Deep Residual Learning for Image Recognition He, Kaiming; Zhang, Xiangyu; Ren, Shaoqing 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) https://doi.org/10.1109/CVPR.2016.90	conference	June 2016
Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inference Jacob, Benoit; Kligys, Skirmantas; Chen, Bo 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) https://doi.org/10.1109/CVPR.2018.00286	conference	June 2018
Machine Learning at Facebook: Understanding Inference at the Edge Wu, Carole-Jean; Brooks, David; Chen, Kevin 2019 IEEE International Symposium on High Performance Computer Architecture (HPCA) https://doi.org/10.1109/HPCA.2019.00048	conference	February 2019
Fast, energy-efficient, robust, and reproducible mixed-signal neuromorphic classifier based on embedded NOR flash memory technology Guo, X.; Bayat, F. Merrikh; Bavandpour, M. 2017 IEEE International Electron Devices Meeting (IEDM) https://doi.org/10.1109/IEDM.2017.8268341	conference	December 2017
Device-aware inference operations in SONOS nonvolatile memory arrays Bennett, Christopher H.; Xiao, T. Patrick; Dellana, Ryan 2020 IEEE International Reliability Physics Symposium (IRPS) https://doi.org/10.1109/IRPS45951.2020.9129313	conference	April 2020
MLPerf Inference Benchmark Reddi, Vijay Janapa; Cheng, Christine; Kanter, David 2020 ACM/IEEE 47th Annual International Symposium on Computer Architecture (ISCA) https://doi.org/10.1109/ISCA45697.2020.00045	conference	May 2020
Multiscale Co-Design Analysis of Energy, Latency, Area, and Accuracy of a ReRAM Analog Neural Training Accelerator Marinella, Matthew J.; Agarwal, Sapan; Hsia, Alexander IEEE Journal on Emerging and Selected Topics in Circuits and Systems, Vol. 8, Issue 1 https://doi.org/10.1109/JETCAS.2018.2796379	journal	March 2018
Using Floating-Gate Memory to Train Ideal Accuracy Neural Networks Agarwal, Sapan; Garland, Diana; Niroula, John IEEE Journal on Exploratory Solid-State Computational Devices and Circuits, Vol. 5, Issue 1 https://doi.org/10.1109/JXCDC.2019.2902409	journal	June 2019
An Analog Neural Network Computing Engine Using CMOS-Compatible Charge-Trap-Transistor (CTT) Du, Yuan; Du, Li; Gu, Xuefeng IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, Vol. 38, Issue 10 https://doi.org/10.1109/TCAD.2018.2859237	journal	October 2019
3-D Stacked Synapse Array Based on Charge-Trap Flash Memory for Implementation of Deep Neural Networks Park, Yu Jeong; Kwon, Hui Tae; Kim, Boram IEEE Transactions on Electron Devices, Vol. 66, Issue 1 https://doi.org/10.1109/TED.2018.2881972	journal	January 2019
Radiation Response of SNOS Nonvolatile Transistors McWhorter, P. J.; Miller, S. L.; Dellin, T. A. IEEE Transactions on Nuclear Science, Vol. 33, Issue 6 https://doi.org/10.1109/TNS.1986.4334615	journal	January 1986
Total ionizing dose effects in MOS oxides and devices Oldham, T. R.; McLean, F. B. IEEE Transactions on Nuclear Science, Vol. 50, Issue 3 https://doi.org/10.1109/TNS.2003.812927	journal	June 2003
Total Dose Radiation Response of NROM-Style SOI Non-Volatile Memory Elements Draper, Bruce; Dockerty, Robert; Shaneyfelt, Marty IEEE Transactions on Nuclear Science, Vol. 55, Issue 6 https://doi.org/10.1109/TNS.2008.2007566	journal	December 2008
Radiation Effects in Flash Memories Gerardin, S.; Bagatin, M.; Paccagnella, A. IEEE Transactions on Nuclear Science, Vol. 60, Issue 3 https://doi.org/10.1109/TNS.2013.2254497	journal	June 2013
Impact of Total Ionizing Dose on the Data Retention of a 65 nm SONOS-Based NOR Flash Puchner, H.; Ruths, P.; Prabhakar, V. IEEE Transactions on Nuclear Science, Vol. 61, Issue 6 https://doi.org/10.1109/TNS.2014.2367451	journal	December 2014
Total Ionizing Dose Effects in 3-D NAND Flash Memories Bagatin, Marta; Gerardin, Simone; Paccagnella, Alessandro IEEE Transactions on Nuclear Science, Vol. 66, Issue 1 https://doi.org/10.1109/TNS.2018.2878911	journal	January 2019
Training a Neural Network on Analog TaO_x ReRAM Devices Irradiated With Heavy Ions: Effects on Classification Accuracy Demonstrated With CrossSim Jacobs-Gedrim, R. B.; Hughart, D. R.; Agarwal, S. IEEE Transactions on Nuclear Science, Vol. 66, Issue 1 https://doi.org/10.1109/TNS.2018.2886229	journal	January 2019
Evaluation of Radiation Effects in RRAM-Based Neuromorphic Computing System for Inference Ye, Zhilu; Liu, Rui; Taggart, Jennifer L. IEEE Transactions on Nuclear Science, Vol. 66, Issue 1 https://doi.org/10.1109/TNS.2018.2886793	journal	January 2019
ISAAC: a convolutional neural network accelerator with in-situ analog arithmetic in crossbars Shafiee, Ali; Nag, Anirban; Muralimanohar, Naveen ACM SIGARCH Computer Architecture News, Vol. 44, Issue 3 https://doi.org/10.1145/3007787.3001139	journal	June 2016
Achieving ideal accuracies in analog neuromorphic computing using periodic carry Agarwal, Sapan; Jacobs Gedrim, Robin B.; Hsia, Alexander H. 2017 Symposium on VLSI Technology https://doi.org/10.23919/VLSIT.2017.7998164	conference	June 2017

Similar Records

An Accurate, Error-Tolerant, and Energy-Efficient Neural Network Inference Engine Based on SONOS Analog Memory

Journal Article · Mon Jan 03 23:00:00 EST 2022 · IEEE Transactions on Circuits and Systems I: Regular Papers · OSTI ID:1842837

New Total-Ionizing-Dose Resistant Data Storing Technique for NAND Flash Memory

Journal Article · Mon Jul 11 00:00:00 EDT 2022 · IEEE Transactions on Device and Materials Reliability · OSTI ID:1882604

Analytical Bit-Error Model of NAND Flash Memories for Dosimetry Application

Journal Article · Fri Nov 12 23:00:00 EST 2021 · IEEE Transactions on Nuclear Science · OSTI ID:1830527

Related Subjects

43 PARTICLE ACCELERATORS
SONOS
charge trap memory
inference accelerators
ionizing radiation
neural networks
neuromorphic computing
total ionizing dose

Ionizing radiation effects in SONOS-based neuromorphic inference accelerators

Citation Formats

References (27)

Similar Records

Related Subjects