吉林大学学报(工学版) ›› 2026, Vol. 56 ›› Issue (8): 2191-2200.doi: 10.13229/j.cnki.jdxbgxb.20250033

• 计算机科学与技术 • 上一篇    

基于少量故障注入样本的GPGPU程序可靠性预测模型构建方法

魏晓辉(),李旭睿,代渊超,于洪梅(),姜楠,岳恒山   

  1. 吉林大学 计算机科学与技术学院，长春 130012
  • 收稿日期:2025-01-10 出版日期:2026-08-01 发布日期:2026-09-02
  • 通讯作者: 于洪梅 E-mail:weixh@jlu.edu.cn;hmyu@jlu.edu.cn
  • 作者简介:魏晓辉（1972-），男，教授，博士 . 研究方向：云计算，高性能计算，容错计算，分布式系统.E-mail：weixh@jlu.edu.cn
  • 基金资助:
    国家重点研发计划项目(2023YFB4502304);国家自然科学基金项目(62302190);国家自然科学基金项目(62272190)

Construction method of reliability prediction model for GPGPU programs based on limited fault injection samples

Xiao-hui WEI(),Xu-rui LI,Yuan-chao DAI,Hong-mei YU(),Nan JIANG,Heng-shan YUE   

  1. College of Computer Science & Technology，Jilin University，Changchun 130012，China
  • Received:2025-01-10 Online:2026-08-01 Published:2026-09-02
  • Contact: Hong-mei YU E-mail:weixh@jlu.edu.cn;hmyu@jlu.edu.cn

RICH HTML

 0   

PDF (PC)

10

摘要：

针对GPGPU程序在软错误扰动下的可靠性预测问题，提出了一种基于少量故障注入样本的预测模型构建方法。通过设计引导规则实现针对性的故障注入样本收集，结合数据增强策略，在少量故障注入样本前提下，实现了高精度的可靠性模型的构建。实验结果表明：本文方法在仅需300个故障注入样本的情况下，预测模型的平均精度达到91.56%。

关键词: 计算机系统结构, 通用图形处理器, 可靠性, 软错误

Abstract:

Reliability prediction of GPGPU programs under soft error perturbations presents significant challenges， particularly when fault injection samples are limited. This paper proposes a novel method for constructing a high-accuracy reliability prediction model using a small number of fault injection samples. The approach introduces guiding rules to optimize fault injection experiments by reducing redundancy and enhancing the quality of collected samples. Furthermore， data augmentation techniques are employed to mitigate the challenges of insufficient training data， enabling effective model learning. Experimental validation on various benchmark programs demonstrates that the proposed method achieves an average prediction accuracy of 91.56% with only 300 fault injection samples. These results highlight the method's capability to reduce experimental overhead while maintaining high prediction performance.

Key words: computer architecture, general-purpose graphics processing units, reliability, soft error

中图分类号: 

  • TP302

图1

GPGPU架构示意图"

图2

线程动态指令数与SDC概率的相似性分析"

图3

故障点位置与SDC概率的分布关系"

表1

启发式特征的Cramér's V相关性检验结果"

启发式特征Cramér's V值
指令操作功能0.339 114
操作数类型0.305 182
程序执行阶段0.364 619
故障点位置0.345 966
故障点比特翻转方向0.282 975

图4

基于少量故障植入样本的GPGPU程序可靠性预测模型构建方法流程图"

图5

集成故障注入样本处理策略对模型性能的影响"

图6

真实与预测SDC比例对比及误差分析"

[1] Li A, Song S L, Wijtvliet M, et al. Sfu-driven transparent approximation acceleration on gpus[C]∥2016 International Conference on Supercomputing, Istanbul, Turkey, 2016: 1-14.
[2] Buck I, Foley T, Horn D, et al. Brook for GPUs: stream computing on graphics hardware[J]. Acm Transactions on Graphics, 2004, 23(3): 777-786.
[3] Zhang Bo-yuan, Tian Jian-nan, Di Sheng, et al. FZ-GPU: A Fast and high-ratio lossy compressor for scientific computing applications on GPUs[C]∥32nd ACM International Symposium on High-Performance Parallel and Distributed Computing, New York, USA, 2023:129-142.
[4] Andrew L, Scott G. Fast algorithms for convolutional neural networks[C]∥2016 IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, NV, USA, 2016: 4013-4021.
[5] Ghorbel A, Ben Amor N, Abid M. GPGPU-Based parallel computing of viola and jones eyes detection algorithm to drive an intelligent wheelchair[J]. Journal of Signal Processing Systems, 2022, 94(12): 1365-1379.
[6] Aiquipa W A, Carpio J D, Garcia J, et al. Analysis of high resolution panchromatic satellite images, based on GPGPU programming[C]∥2nd International Conference on Sensors, Signal and Image Processing, Prague, Czech Republic, 2019: 45-48.
[7] 余淇睿, 张战刚, 李斌, 等. 大气中子及α粒子对芯片软错误的贡献趋势[J]. 电子与封装, 2023, 23(8): 70-76.
Yu Qi-rui, Zhang Zhan-gang, Li Bin, et al. Trends in the contribution of atmospheric neutrons and α particles to the chip soft errors [J]. Electronics and Packaging, 2023, 23(8): 70-76.
[8] Tan J, Goswami N, Li T, et al. Analyzing soft-error vulnerability on GPGPU microarchitecture[C]∥2011 IEEE International Symposium on Workload Characterization, Austin, TX, USA, 2011:226-235.
[9] Wei X, Jiang N, Yue H, et al. ApproxDup: developing an approximate instruction duplication mechanism for efficient SDC detection in GPGPUs[J]. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2024, 43(4): 1051-1064.
[10] Yang L, Nie B, Jog A, et al. Enabling software resilience in GPGPU applications via partial thread protection[C]//43rd IEEE/ACM International Conference on Software Engineering, Madrid, Spain, 2021: 1248-1259.
[11] Li G, Pattabirman K, Cher C Y, et al. Understanding error propagation in GPGPU applications[C]∥2016 International Conference for High Performance Computing, Networking, Storage and Analysis, Salt Lake City, UT, USA, 2016: 240-251.
[12] Fang B, Pattabiraman K, Ripeanu M, et al. GPU-Qin: A methodology for evaluating the error resilience of GPGPU applications[C]∥2014 IEEE International Symposium on Performance Analysis of Systems and Software, Monterey, CA, USA, 2014: 221-230.
[13] Hari S K S, Tsai T, Stephenson M, et al. SASSIFI: an architecture-level fault injection tool for GPU application resilience evaluation[C]∥2017 IEEE International Symposium on Performance Analysis of Systems and Software, Santa Rosa, CA, USA, 2017: 249-258.
[14] Tsai T, Hari S K S, Sullivan M, et al. NVBitFI: dynamic fault injection for GPUs[C]∥51st Annual IEEE/IFIP International Conference on Dependable Systems and Networks, Taipei, China, 2021: 284-291.
[15] 岳恒山. 面向GPGPU体系结构的软错误弹性分析与容错策略研究[D]. 长春: 吉林大学计算机科学与技术学院, 2022.
Yue Heng-shan. Research on soft error resilience analysis and fault-tolerance strategy for GPGPU architecture[D]. Changchun: College of Computer Science and Technology, Jilin University, 2022.
[16] Lindholm E, Nickolls J, Oberman S, et al. NVIDIA Tesla: a unified graphics and computing architecture[J]. IEEE Micro, 2008, 28(2): 39-55.
[17] Sangchoolie B, Pattabiraman K, Karlsson J. One bit is (Not) enough: an empirical study of the impact of single and multiple bit-flip errors[C]∥47th Annual IEEE/IFIP International Conference on Dependable Systems and Networks, Denver, CO, USA, 2017: 97-108.
[18] Anwer A R, Li G, Pattabiraman K, et al. GPU-Trident: efficient modeling of error propagation in gpu programs[C]∥International Conference for High Performance Computing, Networking, Storage and Analysis, Atlanta, Georgia, USA, 2020: 1-15.
[19] 马骏驰, 汪芸, 蔡震波, 等. 基于错误传播分析的SDC脆弱指令识别方法[J]. 计算机研究与发展, 2016, 53(9): 1943-1952.
Ma Jun-chi, Wang Yun, Cai Zhen-bo, et al. An Approach for Identifying SDC-causing instructions by fault propagation analysis[J]. Journal of Computer Research and Development, 2016, 53(9): 1943-1952.
[20] CRAMÉR H. Mathematical methods of statistics[M]. New Jersey: Princeton University Press, 1946.
[21] Bengio Y. Learning deep architectures for AI[J]. Foundations and Trends in Machine Learning, 2009, 2(1): 1-127
[22] Breiman L, Friedman J H, Olshen R A, et al. Classification and Regression Trees[J]. Belmont: Wadsworth, 1984.
[1] 郑智群,黄贤振,王钰平,路至远. 螺栓组连接高精度变形状况及可靠性分析[J]. 吉林大学学报(工学版), 2026, 56(3): 653-661.
[2] 张云翔,宋现敏,谢渝,湛天舒. 基于用户满意度的停车预约服务智能体行为仿真[J]. 吉林大学学报(工学版), 2025, 55(9): 2978-2984.
[3] 黄贤振,马明斐,李超,王旭,荣治明. 机床回转工作台变形全局可靠性灵敏度分析[J]. 吉林大学学报(工学版), 2025, 55(6): 1906-1914.
[4] 赵庶旭,孙治朝,王小龙. 移动边缘计算场景中的动态身份认证协议[J]. 吉林大学学报(工学版), 2025, 55(3): 1050-1060.
[5] 黄贤振,李超,孙超,邱开慧. 基于失效模式和动态贝叶斯网络的数控机床可靠性分析[J]. 吉林大学学报(工学版), 2025, 55(11): 3534-3543.
[6] 李强,张凌羽,孟祥宇. 资源高效的聚类协同联邦学习客户端选择方法[J]. 吉林大学学报(工学版), 2025, 55(10): 3337-3345.
[7] 黄贤振,于瑞,刘慧珍,唐继武. 考虑轴承非线性恢复力的主轴振动可靠性分析[J]. 吉林大学学报(工学版), 2025, 55(1): 116-124.
[8] 郑长江,陶童统,陈志超. 基于流量可调重分配的级联失效模型[J]. 吉林大学学报(工学版), 2024, 54(9): 2441-2450.
[9] 黄贤振,郭斌,姜智元,唐继武. 高速电主轴系统振动特性和精度可靠性分析[J]. 吉林大学学报(工学版), 2024, 54(9): 2432-2440.
[10] 魏晓辉,王晨洋,吴旗,郑新阳,于洪梅,岳恒山. 面向脉动阵列神经网络加速器的软错误近似容错设计[J]. 吉林大学学报(工学版), 2024, 54(6): 1746-1755.
[11] 陈超,戴孟初,周乐,梁云东. 基于改进综合因素法的农业机械可靠性分配[J]. 吉林大学学报(工学版), 2024, 54(5): 1493-1500.
[12] 樊学平,周衡,刘月飞. 基于Gaussian Copula-贝叶斯动态模型的桥梁构件时变可靠性分析[J]. 吉林大学学报(工学版), 2024, 54(2): 485-493.
[13] 余建星,韦明秀,余杨,崔宇朋,潘宇. 加筋板结构可靠性拓扑优化与工程化设计[J]. 吉林大学学报(工学版), 2024, 54(10): 2781-2791.
[14] 黄贤振,孙楷铂,栾晓刚,胡兵. 螺栓预紧连接可靠性灵敏度分析[J]. 吉林大学学报(工学版), 2023, 53(8): 2219-2226.
[15] 白丽丽,姜封国,周玉明,曾枭. 基于改进鲸鱼算法的结构可靠性优化设计[J]. 吉林大学学报(工学版), 2023, 53(11): 3160-3165.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
版权所有 © 《吉林大学学报(工学版)》编辑部
地址：长春市人民大街5988号　电话：+86-431-85095297　传真：+86-431-85094074　E-mail: xbgxb@jlu.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发