And-parallel execution of logic programs on a shared-Memory multiprocessor

摘要: Abstract This paper presents the implementation and performance results of an -parallel execution model logic programs on a shared-memory multiprocessor. The is meant for with “don't-know nondeterminism”, handles binding conflicts by dynamically detecting dependencies among literals. also incorporates intelligent backtracking at clause level. Our this based upon Warren Machine (WAM); hence it retains most efficiency WAM sequential segments programs. Performance Sequent Balance 21000 show that suitable programs, our parallel can achieve linear speedup dozens processors. We present analysis different overheads encountered in model.