Determining a Failure Root Cause Distribution From a Population of Layout-Aware Scan Diagnosis Results

摘要

THE YIELD OF an integrated circuit (IC) is well known to be a critical factor in the success of an IC in the market place. Achieving high stable yields helps ensure that the product is profitable and meets quality and reliability objectives. When a new manufacturing process is introduced, or a new product is introduced on a mature manufacturing process, yields will tend to be significantly lower than acceptable. The ability to meet profitability and quality objectives, and perhaps more importantly, time-to-market and time-to-volume objectives depend greatly on the rate at which these low yields can be ramped up. While the yield ramp depends on both the yield learning and yield enhancement cycle times, this work focuses on significantly increasing the value of test data and the yield learning rate. Several recent works have investigated methods to extract meaningful information about the failure mechanisms causing yield loss from volume scan diagnosis results 1-7. Scan diagnosis uses design information and tester failing cycle data to identify specific locations in the design that are likely to explain these failing cycles. The primary challenge for yield analysis of diagnosis data is dealing with the ambiguity present in the diagnosis results. The ambiguity in this process is twofold. First, in diagnosis, it is often possible that more than one location can explain the defective logical behavior observed in the failing cycles. Second, each suspect location will often have multiple possible root causes associated with it. Many of the aforementioned works have proposed aggregation of the raw diagnosis results, including the ambiguity, and still extract some meaningful information. In practice, these techniques have limited applicability. In contrast, a subset of these works has focused on statistical techniques aimed at eliminating the ambiguity from the aggregation of results. This work fits within this subset.