When the design is complex and margins are tight, there may be too many possible design tradeoffs to explore with interactive modeling and simulation. HyperLynx has multiple automated optimizers that will improve design performance based on specific performance metrics. Different methods are used depending on the application and size of the design space to be explored. For uncomplicated cases, swept-parameter analysis is simple to set up and provides results that are easy to interpret. In intermediate cases, expert-based algorithms are used to select design alternative for exploration based on application-specific knowledge. In the largest, most complex cases, response surface techniques are used, with advanced algorithms that automatically manage the tradeoff between properly sampling a large design space and zeroing in on areas of optimum performance. The output from each optimization process is an improved set of design parameters and rules that can be driven into PCB layout.