mlsanityData to 14 Oct 2026

Evaluating and Enhancing Robustness of Deep Recommendation Systems Against Hardware Errors

arXiv 2307.10244 ↗First posted 17 Jul 2023cs.IR cs.LGcited by 1

Not on the map. Why?

Abstract (arXiv)

Deep recommendation systems (DRS) heavily depend on specialized HPC hardware and accelerators to optimize energy, efficiency, and recommendation quality. Despite the growing number of hardware errors observed in large-scale fleet systems where DRS are deployed, the robustness of DRS has been largely overlooked. This paper presents the first systematic study of DRS robustness against hardware errors. We develop Terrorch, a user-friendly, efficient and flexible error injection framework on top of the widely-used PyTorch. We evaluate a wide range of models and datasets and observe that the DRS robustness against hardware errors is influenced by various factors from model parameters to input characteristics. We also explore 3 error mitigation methods including algorithm based fault tolerance (ABFT), activation clipping and selective bit protection (SBP). We find that applying activation clipping can recover up to 30% of the degraded AUC-ROC score, making it a promising mitigation method.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 24 of its 60 references that are in the map. First posted on arXiv 17 Jul 2023.

Builds on

24 of 60 references.

References · 24

  1. 2017MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applicationscited by 3,845
  2. 2015Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classificationcited by 3,611
  3. 2018Graph Convolutional Neural Networks for Web-Scale Recommender Systemscited by 1,138
  4. 2017Neural Collaborative Filteringcited by 1,111
  5. 2016Wide & Deep Learning for Recommender Systemscited by 747
  6. 2017DeepFM: A Factorization-Machine based Neural Network for CTR Predictioncited by 676
  7. 2018PACT: Parameterized Clipping Activation for Quantized Neural Networkscited by 467
  8. 2017Deep & Cross Network for Ad Click Predictionscited by 393
  9. 2019Deep Learning Recommendation Model for Personalization and Recommendation Systemscited by 327
  10. 2017Attentional Factorization Machines: Learning the Weight of Feature Interactions via Attention Networkscited by 126
  11. 2019The Architectural Implications of Facebook's DNN-based Personalized Recommendationcited by 83
  12. 2020Distributed Hierarchical GPU Parameter Server for Massive Scale Deep Learning Ads Systemscited by 60
12 more
  1. 2020Understanding Training Efficiency of Deep Learning Recommendation Models at Scalecited by 43
  2. 2019FT-ClipAct: Resilience Analysis of Deep Neural Networks and Improving their Fault Tolerance using Clipped Activationcited by 31
  3. 2020Understanding Capacity-Driven Scale-Out Neural Recommendation Inferencecited by 29
  4. 2018Analyzing and Mitigating the Impact of Permanent Faults on a Systolic Array Based Neural Network Acceleratorcited by 25
  5. 2022Heterogeneous Acceleration Pipeline for Recommendation System Trainingcited by 13
  6. 2020Cross-Stack Workload Characterization of Deep Recommendation Systemscited by 11
  7. 2021RecPipe: Co-designing Models and Hardware to Jointly Optimize Recommendation Quality and Performancecited by 8
  8. 2022Detecting silent data corruptions in the wildcited by 7
  9. 2021Efficient Soft-Error Detection for Low-precision Deep Learning Recommendation Modelscited by 4
  10. 2021MoRS: An Approximate Fault Modelling Framework for Reduced-Voltage SRAMscited by 2
  11. 2022Understanding RowHammer Under Reduced Wordline Voltage: An Experimental Study Using Real DRAM Devicescited by 2
  12. 2022Thales: Formulating and Estimating Architectural Vulnerability Factors for DNN Acceleratorscited by 2

Built on by

Cited by year

Citing papers per year: 2025: 1.

1 paper.

  1. 2025Understanding Silent Data Corruption in LLM Trainingcited by 15