Optimal testing strategy in semiconductor testing process

Abstract

We present an optimal test resource allocation strategy using uncertainty reduction in an environment where resource capacity changes dynamically according to engineering activity. The dynamics of test capacity change are modeled using a linear programming model and then extended and generalized to a Markov decision process. We analyze the model to develop structural results and illustrate its behavior with numerical examples. To the best of our knowledge, this model is the first to define, formalize, and analyze the decision-making process associated with reducing final test time in an environment where capacity may be dynamically increased, depending on engineering activity results.