{"ID":2873250,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2509.06289","arxiv_id":"2509.06289","title":"A Spatio-Temporal Graph Neural Networks Approach for Predicting Silent Data Corruption inducing Circuit-Level Faults","abstract":"Silent Data Errors (SDEs) from time-zero defects and aging degrade safety-critical systems. Functional testing detects SDE-related faults but is expensive to simulate. We present a unified spatio-temporal graph convolutional network (ST-GCN) for fast, accurate prediction of long-cycle fault impact probabilities (FIPs) in large sequential circuits, supporting quantitative risk assessment. Gate-level netlists are modeled as spatio-temporal graphs to capture topology and signal timing; dedicated spatial and temporal encoders predict multi-cycle FIPs efficiently. On ISCAS-89 benchmarks, the method reduces simulation time by more than 10x while maintaining high accuracy (mean absolute error 0.024 for 5-cycle predictions). The framework accepts features from testability metrics or fault simulation, allowing efficiency-accuracy trade-offs. A test-point selection study shows that choosing observation points by predicted FIPs improves detection of long-cycle, hard-to-detect faults. The approach scales to SoC-level test strategy optimization and fits downstream electronic design automation flows.","short_abstract":"Silent Data Errors (SDEs) from time-zero defects and aging degrade safety-critical systems. Functional testing detects SDE-related faults but is expensive to simulate. We present a unified spatio-temporal graph convolutional network (ST-GCN) for fast, accurate prediction of long-cycle fault impact probabilities (FIPs)...","url_abs":"https://arxiv.org/abs/2509.06289","url_pdf":"https://arxiv.org/pdf/2509.06289v1","authors":"[\"Shaoqi Wei\",\"Senling Wang\",\"Hiroshi Kai\",\"Yoshinobu Higami\",\"Ruijun Ma\",\"Tianming Ni\",\"Xiaoqing Wen\",\"Hiroshi Takahashi\"]","published":"2025-09-08T02:23:51Z","proceeding":"cs.LG","tasks":"[\"cs.LG\",\"cs.AR\",\"cs.ET\"]","methods":"[\"Graph Neural Network\"]","has_code":false}