Applications of the Taguchi method for optimized package design

A robust technique for noise and sensitivity analysis for packages in the system environment using the design of experiment (DOE) and orthogonal matrix array is presented. The use of DOE allows full factorial parametric analysis and the use of an orthogonal array significantly reduces the number of simulations. The results of noise and sensitivity analysis can be used to predict the noise and sensitivity of system performance to random package physical design parameter variations, to identify the most noise sensitive parameters and to quantify the effects. System performance predictions and sensitivity analysis can be used to compare the performance of different packages, to determine the performance limits of the package and system, and to optimize system and package design. The reliability of this approach is verified by case studies