PERIODIC ADAPTIVE BRANCH PREDICTION AND ITS APPLICATION IN SUPERSCALAR PROCESSING IN PROLOG

Branch instructions create barriers to instruction prefetching, greatl y reducing the fine-grained parallelism of programs. Branch prediction is a common method for solving this problem. We first present four le mmata in this paper describing the relationships among branch predicti on hit rate and system performance, hardware efficiency, and branch pr ediction overhead. We then propose a branch prediction method called P AM (Periodic Adaptive Method). An abstract model and detailed implemen tation of PAM are described. PAM's prediction hit rate as measured by 10 Prolog benchmark programs is 97%. When implemented in a superscalar Prolog system, PAM enhances the degree of system parallelism by 68.8% . PAM can be applied to languages and applications other then the Prol og system we used in this study.