Analytical modeling is applied to the automated design of application-specific superscalar processors. Using an analytical method bridges the gap between the size of the design space and the time required for detailed cycle-accurate simulations. The proposed design framework takes as inputs the design targets (upper bounds on execution time, area, and energy), design alternatives, and one or more application programs. The output is the set of out-of-order superscalar processors that are

more »

... ptimal with respect to performance-energy-area. The core of the new design framework is made up of analytical performance and energy activity models, and an analytical model-based design optimization process. For a set of benchmark programs and a design space of 2000 designs, the design framework arrives at all performanceenergy-area Pareto-optimal design points within 16 minutes on a 2 GHz Pentium-4. In contrast, it is estimated that a naïve cycle-accurate simulation-based exhaustive search would require at least two months to arrive at the Pareto-optimal design points for the same design space. The design optimization process employs a simple divide-and-conquer approach to arrive at optimal designs from the available components. The performance, energy, and area of each configuration are evaluated with the respective models, and a set of Pareto-optimal configurations is produced. From this set of Pareto-optimal configurations the ones that satisfy the design constraints are presented to the system designer. Or, if no designs satisfy the constraints, this information is also conveyed to the designer. For a set of benchmark programs, the design framework takes about 16 minutes to generate a set of designs that are performance-energy-area Pareto-optimal.