An Application of the Linear Errors-in-Variables Model in Semiconductor Device Performance Assessment

Quality Engineering vol. 27 issue 4 - October 2015

Abstract: Designers and manufacturers in the semiconductor industry constantly seek ways of improving device performance without an undue increase in power consumption. A particular concern for battery powered mobile electronics is the increased propensity for current leakage as devices get smaller in size. As a result, assessment of device performance involves joint analysis of measurement data on saturation drain current (a measure of performance) and off-state current (leakage). Both engineering and statistical considerations suggest the linear errors-in-variables (EIV) model as an appropriate model in this context. Moreover, in applications the purpose is often to compare device performance under different design and processing conditions. The EIV model with an indicator variable would provide a suitable framework for such comparative assessments. Even though modeling with indicator variables is well established in ordinary linear regression, the literature and software offerings on the EIV model with an indicator variable are highly sparse. In this article, we extend the traditional EIV model to include an indicator variable in order to enable statistically based comparisons between treatment groups of interest. We use maximum likelihood for model estimation. A real-life case study is used to illustrate the application of the extended model for a comparative analysis of the semiconductor device performance.

Keywords: Case studies; Semiconductor industry; Applications; Assessment; Regression; Confidence intervals;

Access this article
Other ways to access this article
You will need Adobe Reader to view this PDF document. Download the free Reader from Adobe

Featured advertisers

Article
Rating

(0) Member Reviews

Featured advertisers





ASQ is a global community of people passionate about quality, who use the tools, their ideas and expertise to make our world work better. ASQ: The Global Voice of Quality.