Advanced Method for Predicting Lifetime of NAND Memory Data Retention Using Long term (1 Year) Characterization

In the actual storage situation of NAND solution products, the power managing and periodic read operations occur before data storage. Since these operations cause various situation of initial fail bits, it was considered to include these operations in the low temperature data retention (LTDR) evaluation to provide accurate lifetime prediction. An extrapolate method using the power-law relationship is used to predict LTDR lifetime. To increase the accuracy of this method, a calculation method for classifying the characteristics of the initial fail bit of NAND has been proposed.