Exploring test space with fuzzy decision making

h E COMBINATORIAL explosion of design alternatives constitutes the fundamental problem in VU1 design. Each possibility requires consideration to produce a design that meets input constraints and is nearly optimal with respect to d e sign goals. To cope with the increasing number of new designs, the next generation of VLSI design systems will need mechanisms for modeling and managing the design space. Efficient management of d e sign alternatives eases a designer’s search for an optimal solution for a given design task. Additionally, it proves crucial to the maintenance and reuse of existing design versions and design databases. System designers must structure the design process so that the solution space is manageable and d e sign alternatives receive proper evaluation. As a result, the decision-making aspect-the selection of a solution from a number of alternatives according to a set of cost/ performance criteria-will dominate both the design process and time. It b e comes evident that we cannot avoid automating the decision-making process