Parallel Execution of Constraint Handling Rules on a Graphical Processing Unit
Amira I. Zaki, Thom Frühwirth, Ilvar Geller · 2012
Abstract. Graphical Processing Units (GPUs) consist of hundreds of small cores, collectively operating to provide massive computation capabilities. The aim of this work is to utilize this technology to execute Constraint Handling Rules (CHR) which are inherently parallel. A translation scheme is defined to transform a subset of CHR rules to C/C++, then to use a GPU to fire the rules on all combinations of constraints. As proof of concept, the scheme was performed on several CHR examples.