考慮實體資訊及單元資訊之多重缺陷診斷

Physical-aware and Cell-aware Diagnosis of Multiple Defects

李其霖(Chi-Lin Lee)

指導教授 ： 李建模

國立臺灣大學/電機資訊學院/電子工程學研究所/碩士(2017年)

https://doi.org/10.6342/NTU201702294

摘要

這篇論文針對多重缺陷的錯誤晶片提出一個考慮實體資訊及單元資訊之診斷技術，而且我們的技術考慮拜占庭效應。我們使用一種數據挖掘的技術將可能的缺陷候選組成叢聚以處理多重缺陷。在ISCAS’89和ITC’99的電路上，多重缺陷的實驗顯示出我們診斷技術的能力。我們的診斷技術有著比商業軟體還要高的正確率，尤其是當插入的缺陷數量越多，差距越大。當插入十個多重導線開路缺陷時，此技術的正確率(0.72)比商業軟體的正確率(0.31)還要高。當插入十個多重橋接缺陷時，我們的診斷技術的正確率(0.70)比商業軟體的正確率(0.26)還要高得更多。而當插入十個單元內缺陷時，我們的技術可以達到0.70的正確率。與我們之前的技術相比[Chen 15]，我們的執行時間改善了79%。

關鍵字

多重缺陷診斷 ； 考慮實體資訊 ； 考慮單元資訊

並列摘要

This thesis presents a physical-aware and cell-aware diagnosis technique for failing dies with multiple defects. Our diagnosis technique considers Byzantine effect. We use a data mining technique to group defect candidates into clusters to handle multiple defects. Simulations on ISCAS’89 and ITC’99 benchmark circuits with multiple defects demonstrate the effectiveness of our diagnosis technique. The accuracy of our technique is higher than commercial tool, especially much higher when more defects are injected. The accuracy with ten open defects injected (0.72) is higher than commercial tool (0.31). The accuracy with ten bridging defects injected (0.70) is much higher than commercial tool (0.26). And for intra-cell defects, our technique can achieve accuracy of 0.70 when ten defects are injected. Our proposed technique achieves 79% runtime reduction compared to our old algorithm [Chen 15].

並列關鍵字

Multiple defect diagnosis ； Physical-aware ； Cell-aware

參考文獻

[Abramovici 84] M. Abramovici, P. R. Menon, and D. T. Miller, “Critical Path Tracing: An Alternative to Fault Simulation,” in Proc. of Design Automation Conference, 1984.

[Acken 92] J. M. Acken, and S. D. Millman, "Fault Model Evolution for Diagnosis: Accuracy vs Precision," in Proc. of Custom Integrated Circuits Conference, 1992.

[Amyeen 06] E. Amyeen, D. Nayak, and S. Venkataraman, “Improving Precision Using Mixed-level Fault Diagnosis,” in Proc. of International Test Conference, 2006.

[Bartenstein 01] T. Bartenstein, D. Heaberlin, L. Huisman, and D. Sliwinski, “Diagnosing Combinational Logic Designs Using the Single Location At-a-time (SLAT) Paradigm,” in Proc. of International Test Conference, 2001.

[Chao 14] S.-M. Chao, P.-J. Chen, and James C.-M. Li, “Divide and Conquer Diagnosis for Multiple Defects,” in Proc. of International Test Conference, 2014.