Value driver derivations for embedded memories by fuzzy DEMATEL based hybrid MCDM methods

The embedded memories have already played a very important role in the modern system on a chip (SoC) designs from the aspects of performance enhancement, power reduction, form factor optimization, etc. Meanwhile, embedded memories usually occupied for the major portion of an SoC layout. Thus, how an embedded memory IP (intellectual property) provider can differentiate the product to maximize the value of that IP has already become a key factor for profit maximization. However, very few scholars tried to resolve such problems. The product and service differentiation problem of memory IPs is in nature a multiple attribute decision making (MADM) problem. Moreover, the differentiators of an embedded memory IP usually interact with each other. Furthermore, evaluations of customers' needs and wants, which are vague, are often required for such a product marketing strategy definition problem. Thus, a fuzzy DEMATEL based multiple criteria decision making (MCDM) framework will be proposed for maximizing the value of an embedded memory IP. The value driver for differentiating an embedded memory IP will first be derived by literature review. The most influential value drivers will be selected. Then, strategies will be proposed for maximizing the value through the most influential value drivers. An empirical study based on an embedded memory IP being provided by an SOC design service provider will be used for verifying the feasibility of the framework. Based on the empirical study results, both power consumption as well as the area of the embedded memory core are the most important value drivers. The proposed product marketing strategies can also be used for maximizing the value of the embedded IP.