Towards Improvement of Mission Mode Failure Diagnosis for System-on-Chip

Abstract

In critical (e.g. automotive) applications, Systems-on-Chip (SoC) failures that occurred during mission mode (in the field) are the most critical since they may lead to catastrophic effects. In this context, diagnosis is crucial in order to establish the root cause of observed failures with the best accuracy. With the advent of very deep submicron technologies (i.e. 7 nm), achieving such level of accuracy will become more and more difficult with today's intra-cell diagnosis tools based on effectcause or cause-effect paradigms. This will compromise the success of subsequent Physical Failure Analysis (PFA) done on defective SoCs. Machine Learning (ML) is now used in numerous classification problems where the knowledge on some data can be used to classify a new instance of such data. In particular, several ML-based solutions exist to address volume diagnosis for yield improvement. These learning-guided diagnosis approaches start from an existing set of defect candidates and try to minimize this set (eliminate bad candidates) owing to the use of ML tools and numerous data collected during production test (e.g. thousands of failed chips with candidates correctly labeled). Although efficient in volume diagnosis, these approaches cannot be used to identify the root cause of failures in customer returns, since only one failed chip is investigated in this case, with no information about the defective behavior of some other similar chips used in the same conditions (environment, workload, etc.). In this paper, we propose a new learning-guided approach for diagnosis of mission mode failures in customer returns. The proposed approach directly produces a minimum set of good candidates derived from the application of the learning-guided intra-cell diagnosis flow. Results obtained on a set of benchmark circuits, and comparison with a commercial intra-cell diagnosis tool, show the feasibility, effectiveness and accuracy of the proposed approach.

Cell-Aware Defect Diagnosis of Customer Returns Based on Supervised Learning

“…Several supervised learning algorithms were considered, with various levels of efficacy. Results obtained on combinational benchmark circuits, and comparison with a commercial cellaware diagnosis tool, show the feasibility and accuracy of this approach [4]. In the second part of the paper, the previous work has been extended by dealing with more sophisticated (i.e.…”

“…In this section, we present a new approach that uses supervised learning instead of traditional cause-effect and/or effect-cause analysis to identify static defect candidates within a cell with a high accuracy. Figure 1 shows the proposed diagnosis flow [4] based on supervised learning that takes a known set of input data and known responses (labeled data) used as training data, trains a model, and then implement a classifier based on this model to make predictions (inferences) for the response to new data.…”

“…This time, we concentrated on a Bayesian classification method for predicting the nature (likelihood to be a good candidate) of each new data instance (defect) that has to be evaluated. This choice comes from the results obtained in [4] after experimenting several learning algorithms and observing their prediction accuracies. These results are partially reported again in Section VI.A.…”

Learning-Based Cell-Aware Defect Diagnosis of Customer Returns

“…temperature and voltage) that will reveal the failure. In case of latent defect, the task will often succeed and a diagnosis program made of several routines is used to identify, step by step, the failing part and, finally, the suspected defects [2]. Each routine corresponds to the application of a diagnosis algorithm at a given hierarchy level.…”

“…As a result, many efforts have been dedicated recently for improving resolution by using machine learning techniques, primarily through the derivation of characteristics that enables correct candidates (candidates that correctly represent defect locations) to be distinguished from incorrect ones [6]- [8]. Though efficient, a common feature of these techniques is that they all address volume diagnosis for yield improvement, which is a different problem than fault diagnosis of customer returns [2]. Indeed, during volume diagnosis, numerous data collected during manufacturing test and subsequent diagnosis phases are available, such as, e.g.…”