Abstract: It is necessary to compute the process capability index (PCI) of non-normal data when considering the level of semiconductor processes.Several main PCIs that have already been presented are analyzed,and their advantages and disadvantages are presented.Based on Chebyshev-Hermite polynomials,a model for computing non-normal PCIs is then given when regarding the fact that these four moments,i.e.mean,standard deviation,skewness,and kurtosis,are suitable to approximately characterize the data distribution properties,which is effective when data deviation is large.Finally,an example is given for proving the relational model.

Key words: non-normal distribution, process capability index, Chebyshev-Hermite polynomials, quality control