利用子集關係進行系統性延遲錯誤診斷

Abstract

由於技術節點的減小和操作頻率的增加，延遲故障在現代設計中變得越來越重要。然而，與時序相關的測試失敗通常只有很少的故障位，因此延遲故障的診斷是具有挑戰性的。本研究提出了一種兩階段流程，用於識別系統性延遲故障並提高相對應的診斷解析度。在第一階段，從延遲故障中觀察到大量測試失敗之間的子集關係。這些子集關係被進一步分析以識別系統性缺陷。在第二階段，選擇一組測試失敗作為代表性測試失敗，用於表示缺陷行為。這些代表性測試失敗被用來做進一步的診斷。將提出的技術實驗在一個工業設計晶片中，其中兩個核心上顯示具有三個系統性案例，透過所提出的技術，延遲故障診斷解析度分別提高了超過 33 倍、69 倍和 8 倍。此外，所提出的技術可以融合進現有容量診斷方法，並可在執行過程中與當前商業診斷工具進行整合。

Delay faults have become increasingly important in modern designs due to decreasing technology node size and increasing operation frequency. However, diagnosis of delay faults can be challenging since there are typically few failing bits in the timing-related test failures. In this work, a two-phase flow is presented to identify systematic delay failures and improve their corresponding diagnosis resolution. In the first phase, a significant amount of subset relationships among test failures are observed for the delay failures. These subset relationships are further analyzed to identify systematic defects. In the second phase, a set of test failures are selected as representative test failures, which are used to represent the defect behavior. These representative test failures are further applied to diagnosis. Experiments on two cores of an industrial design with three systematic cases show over 33×, 69×, and 8× improvement on delay fault diagnosis resolution. Furthermore, the proposed technique can integrate with existing volume diagnosis methodologies, and incorporates current commercial diagnosis tools in its execution.