Design and implementation of a processor core with a double-instruction pipeline and no need of branch prediction