Test-data volume optimization for diagnosis

Abstract

Test data collection for a failing integrated circuit (IC) can be very expensive and time consuming. Many companies now collect a fix amount of test data regardless of the failure characteristics. As a result, limited data collection could lead to inaccurate diagnosis, while an excessive amount increases the cost not only in terms of unnecessary test data collection but also increased cost for test execution and data-storage. In this work, the objective is to develop a method for predicting the precise amount of test data necessary to produce an accurate diagnosis. By analyzing the failing outputs of an IC during its actual test, the developed method dynamically determines which failing test pattern to terminate testing, producing an amount of test data that is sufficient for an accurate diagnosis analysis. The method leverages several statistical learning techniques, and is evaluated using actual data from a population of failing chips and five standard benchmarks. Experiments demonstrate that test-data collection can be reduced by > 30% (as compared to collecting the full-failure response) while at the same time ensuring >90% diagnosis accuracy. Prematurely terminating test-data collection at fixed levels (e.g., 100 failing bits) is also shown to negatively impact diagnosis accuracy.

An Expert CAD Flow for Incremental Functional Diagnosis of Complex Electronic Boards

“…A comparative analysis of artificial intelligence-based techniques for extracting the system model from test data volumes is presented in [6], evaluating the amount of data needed by the different methods with respect to the achieved accuracy of the final diagnosis, when considering complete syndromes.…”

“…In the latter case, an extensive research has been performed to identify the most effective machine learning technique, spanning from Bayesian inference [2], to decision trees (DTs) [3], from artificial neural networks (ANNs) to support vector machines [4], [5]. A comparative analysis to evaluate the effectiveness of these techniques in extracting information from test-data volumes is proposed in [6]. Indeed it is important to be able to extract the knowledge from the previous historical data; however, we here focus on the subsequent diagnostic process, and its efficiency, in terms of the amount of information (test results) needed to identify and isolate the candidate faulty component.…”

A configurable board-level adaptive incremental diagnosis technique based on decision trees

“…Much research has been devoted to the extraction of an effective model of the system from a limited amount of historic data [10], [11]. In this work we do not address the problem of incomplete or insufficient information, focusing instead on the other alternative for building the decision tree, starting from a model provided by the test/diagnosis engineer, as the one used by commercial tools [12], [13] typically adopted in industrial environments.…”

Improving the accuracy of defect diagnosis by considering reduced diagnostic information

“…This is the basis for the computation of small fault dictionaries [21], for the reduction of test data volume for defect diagnosis [22], This work was supported in part by SRC Grant No. 20J3-TJ-2469.…”