{CUDA}: Set constraints on GPUs

Abstract

Set constraints have been introduced in declarative programming languages in the Nineties as a consequence of a broader research on programming with sets and on computable set theory. General Purpose Graphics Processing Units (GPUs), originally developed for graphical purposes (e.g., for high definition video games), emerged recently as a powerful and cheap parallel architecture, widely available in most desktops and laptops computers. This paper presents a constraint solver on set constraints and its parallel implementation on GPUs.