Single event effects evaluation on convolution neural network in Xilinx 28 nm system on chip

doi:10.1088/1674-1056/ad3b82

中国物理B ›› 2024, Vol. 33 ›› Issue (7): 78501-078501.doi: 10.1088/1674-1056/ad3b82

Single event effects evaluation on convolution neural network in Xilinx 28 nm system on chip

Xu Zhao(赵旭)¹, Xuecheng Du(杜雪成)^1,†, Xu Xiong(熊旭)¹, Chao Ma(马超)¹, Weitao Yang(杨卫涛)², Bo Zheng(郑波)¹, and Chao Zhou(周超)¹

1 School of Nuclear Science and Technology, University of South China, Hengyang 421001, China;
2 School of Microelectronics, Xidian University, Xi'an 710071, China

收稿日期:2024-01-09 修回日期:2024-02-29 接受日期:2024-04-07 出版日期:2024-06-18 发布日期:2024-06-18
通讯作者: Xuecheng Du E-mail:duxuecheng@usc.edu.cn
基金资助:
Project supported by the National Natural Science Foundation of China (Grant No. 12305303), the Natural Science Foundation of Hunan Province of China (Grant Nos. 2023JJ40520, 2021JJ40444, and 2019JJ30019), the Research Foundation of Education Bureau of Hunan Province of China (Grant No. 20A430), the Science and Technology Innovation Program of Hunan Province (Grant No. 2020RC3054), the Natural Science Basic Research Plan in the Shaanxi Province of China (Grant No. 2023-JC-QN-0015), and the Doctoral Research Fund of University of South China.

Single event effects evaluation on convolution neural network in Xilinx 28 nm system on chip

Xu Zhao(赵旭)¹, Xuecheng Du(杜雪成)^1,†, Xu Xiong(熊旭)¹, Chao Ma(马超)¹, Weitao Yang(杨卫涛)², Bo Zheng(郑波)¹, and Chao Zhou(周超)¹

1 School of Nuclear Science and Technology, University of South China, Hengyang 421001, China;
2 School of Microelectronics, Xidian University, Xi'an 710071, China

Received:2024-01-09 Revised:2024-02-29 Accepted:2024-04-07 Online:2024-06-18 Published:2024-06-18
Contact: Xuecheng Du E-mail:duxuecheng@usc.edu.cn
Supported by:
Project supported by the National Natural Science Foundation of China (Grant No. 12305303), the Natural Science Foundation of Hunan Province of China (Grant Nos. 2023JJ40520, 2021JJ40444, and 2019JJ30019), the Research Foundation of Education Bureau of Hunan Province of China (Grant No. 20A430), the Science and Technology Innovation Program of Hunan Province (Grant No. 2020RC3054), the Natural Science Basic Research Plan in the Shaanxi Province of China (Grant No. 2023-JC-QN-0015), and the Doctoral Research Fund of University of South China.

摘要/Abstract

摘要： Convolutional neural networks (CNNs) exhibit excellent performance in the areas of image recognition and object detection, which can enhance the intelligence level of spacecraft. However, in aerospace, energetic particles, such as heavy ions, protons, and alpha particles, can induce single event effects (SEEs) that lead CNNs to malfunction and can significantly impact the reliability of a CNN system. In this paper, the MNIST CNN system was constructed based on a 28 nm system-on-chip (SoC), and then an alpha particle irradiation experiment and fault injection were applied to evaluate the SEE of the CNN system. Various types of soft errors in the CNN system have been detected, and the SEE cross sections have been calculated. Furthermore, the mechanisms behind some soft errors have been explained. This research will provide technical support for the design of radiation-resistant artificial intelligence chips.

关键词: single event effects, convolutional neural networks, alpha particle, system on chip, fault injection

Abstract: Convolutional neural networks (CNNs) exhibit excellent performance in the areas of image recognition and object detection, which can enhance the intelligence level of spacecraft. However, in aerospace, energetic particles, such as heavy ions, protons, and alpha particles, can induce single event effects (SEEs) that lead CNNs to malfunction and can significantly impact the reliability of a CNN system. In this paper, the MNIST CNN system was constructed based on a 28 nm system-on-chip (SoC), and then an alpha particle irradiation experiment and fault injection were applied to evaluate the SEE of the CNN system. Various types of soft errors in the CNN system have been detected, and the SEE cross sections have been calculated. Furthermore, the mechanisms behind some soft errors have been explained. This research will provide technical support for the design of radiation-resistant artificial intelligence chips.

Key words: single event effects, convolutional neural networks, alpha particle, system on chip, fault injection

中图分类号: (Semiconductor-device characterization, design, and modeling)

85.30.De

61.82.Fk (Semiconductors) 85.35.-p (Nanoelectronic devices)

Xu Zhao(赵旭), Xuecheng Du(杜雪成), Xu Xiong(熊旭), Chao Ma(马超), Weitao Yang(杨卫涛), Bo Zheng(郑波), and Chao Zhou(周超). Single event effects evaluation on convolution neural network in Xilinx 28 nm system on chip[J]. 中国物理B, 2024, 33(7): 78501-078501.

参考文献

[1] Zhao G, Hu J Y, Xiao W L and Zou J 2021 Chin. J. Aeronaut. 34 214
[2] Cosmas K and Kenichi A 2020 Aerospace 7 159
[3] Arena P, Basile A, Bucolo M and Fortuna L 2003 Nucl. Instrum. Meth. A 497 174
[4] Guo Y, Chen W and Jing S W 2020 Chin. Phys. B 29 054203
[5] Ma Y F, Cao Y, Vrudhula S and Seo J S 2018 IEEE Trans. Very Large Scale Integr VLSI Syst. 26 1354
[6] Wei X H, Zhou C B, Yue H S and Zhou J T Y 2023 J. Syst. Archit. 145 103024
[7] Libano F, Wilson B, Wirthlin M, Rech P and Brunhaver J 2020 IEEE Trans. Nucl. Sci. 67 1478
[8] Autran J L, Munteanu D, Roche P, Gasiot G, Martinie S, Uznanski S, Sauze S, Semikh S, Yakushev E, Rozov S, Loaiza P, Warot G and Zampaolo M 2010 Microelectron Reliab. 50 1822
[9] Baumann R C 2005 IEEE T. Device Mat. RE. 5 305
[10] Yang W T, Li Y H, Guo Y X, Zhao H Y, Li Y, Li P, He C H, Guo G, Liu J, Yang S S and An H 2020 Chin. Phys. B 29 108504
[11] Jin X M, Chen W, Li J L, Qi C, Guo X Q, Li R B and Liu Y 2019 Chin. Phys. B 28 104212
[12] Xilinx, lnc. 2023 Soft Error Mitigation Controller LogiCORE IP Product Guide, PG036 (v4.1)
[13] Toner C, Boukabache H, Ducos G, Pangallo M, Danzeca S, Widorski M, Roesler S and Perrin D 2019 Microelectron Reliab. 100 113492
[14] Yang W T, Du B Y, He C H and Sterpone L 2021 Microelectron Reliab. 120 114122
[15] Li W S, Ge G J, Guo K Y, Chen X M, Wei Q, Gao Z, Wang Y and Yang H Z 2020 2nd IEEE International Conference on Artificial Intelligence Circuits and Systems (AICAS), August 31-September 02, 2020, Genova, Italy, p. 1
[16] Gao Z, Gao S H, Yao Y, Liu Q, Zeng S L, Ge G J, Wang Y, Ullah A and Reviriego P 2023 IEEE Trans. Device Mater Reliab. 23 116
[17] Wang H B, Wang Y S, Xiao J H, Wang S L and Liang T J 2021 IEEE Trans. Nucl. Sci. 68 394
[18] Alzubaidi L, Zhang J L, Humaidi A J, AI-Dujaili A, Duan Y, AIShamma O, Santamaría J, A.Fadhel M, AI-Amidie M and Farhan L 2021 J. Big. Data 8 1
[19] dos Santos A F, Tambara L A, Benevenuti F, Tonfat J and Kastensmidt F L 2017 13th Interna-tional Symposium on Applied Reconfigurable Computing (ARC), April 3-7, 2017, Delft, Netherlands, p. 202
[20] Zhang Z G, Ye B, Ji Q G, Guo J L, Xi K, Lei Z F, Huang Y, Peng C, He Y J, Liu J and Du G H 2020 Acta Phys. Sin. 69 136103 (in Chinese)
[21] JESD89 A Measurement and Reporting of Alpha Particle and Terrestrial Cosmic Ray-Induced Soft Errors in Semiconductor Devices JEDEC standard, October 2006
[22] Howard J W, Carts M A, Stattel R, Rogers C E, Lrwin T L, Dunsmore C, Sciarini J A and LaBel K A 2001 IEEE Radiation Effects Data Workshop (REDW) and IEEE Nuclear and Space Ra-diation Effects Conference (NSREC), July 16-20, 2001, Vancouver, BC, Canada, p. 38
[23] Hsueh M C, Tsai T K and Iyer R K 1997 Computer 30 75
[24] Wang Z B, Chen W, Yao Z B, Zhang F Q, Luo Y H, Tang X B, Peng C, Ding L L and Guo X Q 2020 Nucl. Instrum. Meth. A 966 163866
[25] Xilinx, lnc. 2012 Soft error mitigation using prioritized essential bits XAPP538(v1.0)

Single event effects evaluation on convolution neural network in Xilinx 28 nm system on chip

Single event effects evaluation on convolution neural network in Xilinx 28 nm system on chip

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 9

Metrics

本文评价

推荐阅读 10

[1]	Niuqi Li(李钮琦), Yingfeng Xu(徐颖峰), Fangchuan Zhong(钟方川), and Debing Zhang(张德兵). Numerical study of alpha particle loss with toroidal field ripple based on CFETR steady-state scenario[J]. 中国物理B, 2024, 33(1): 15202-15202.
[2]	Qun-Si Yang(羊群思), Qing Liu(刘清), Dong Zhou(周东), Wei-Zong Xu(徐尉宗), Yi-Wang Wang(王宜望), Fang-Fang Ren(任芳芳), and Hai Lu(陆海). Alpha particle detector with planar double Schottky contacts directly fabricated on semi-insulating GaN:Fe template[J]. 中国物理B, 2021, 30(11): 117303-117303.
[3]	叶鑫, 夏晓川, 梁红伟, 李卓, 张贺秋, 杜国同, 崔兴柱, 梁晓华. Effect of Au/Ni/4H-SiC Schottky junction thermal stability on performance of alpha particle detection[J]. 中国物理B, 2018, 27(8): 87304-087304.
[4]	雷志锋, 张战刚, 恩云飞, 黄云. Mechanisms of atmospheric neutron-induced single event upsets in nanometric SOI and bulk SRAM devices based on experiment-verified simulation tool[J]. 中国物理B, 2018, 27(6): 66105-066105.
[5]	张晋新, 郭红霞, 潘霄宇, 郭旗, 张凤祁, 冯娟, 王信, 魏莹, 吴宪祥. Synergistic effect of total ionizing dose on single event effect induced by pulsed laser microbeam on SiGe heterojunction bipolar transistor[J]. 中国物理B, 2018, 27(10): 108501-108501.
[6]	张晋新, 贺朝会, 郭红霞, 李培, 郭宝龙, 吴宪祥. Three-dimensional simulation of fabrication process-dependent effects on single event effects of SiGe heterojunction bipolar transistor[J]. 中国物理B, 2017, 26(8): 88502-088502.
[7]	张战刚, 刘杰, 侯明东, 孙友梅, 赵发展, 刘刚, 韩郑生, 耿超, 刘建德, 习凯, 段敬来, 姚会军, 莫丹, 罗捷, 古松, 刘天奇. Large energy-loss straggling of swift heavy ions in ultra-thin active silicon layers[J]. 中国物理B, 2013, 22(9): 96103-096103.
[8]	张战刚, 刘杰, 侯明东, 孙友梅, 苏弘, 段敬来, 莫丹, 姚会军, 罗捷, 古松, 耿超, 习凯. Angular dependence of multiple-bit upset response in static random access memories under heavy ion irradiation[J]. 中国物理B, 2013, 22(8): 86102-086102.
[9]	高博, 刘刚, 王立新, 韩郑生, 宋李梅, 张彦飞, 腾瑞, 吴海舟. Radiation damage effects on power VDMOS devices with composite SiO₂–Si₃N₄ films[J]. 中国物理B, 2013, 22(3): 36103-036103.