Automatic analysis of memory faulty behavior in defective memories

As the complexity of memory faulty behavior increases, it is becoming more difficult to precisely identify the faults the memory exhibits. Knowledge of the precise set of faults is essential for designing an optimal set of memory tests with low test time and high fault coverage. This paper presents an automatic method to analyze the observed faulty behavior and to map it precisely into corresponding faults. The method is unique in its generality, making it possible to identify both static as well as dynamic faults in the behavior. Depending on the complexity of the performed fault analysis, different algorithms may be used, with increasing level of computational complexity.