ABSTRACT
In this book chapter, we describe Graph Engine for Multithreaded Systems (GEMS), a full soware stack that implements a graph database on a commodity cluster and enables scaling in data set size while maintaining a constant query throughput when adding more cluster nodes. GEMS employs the SPARQL (SPARQL Protocol and RDF Query Language) language for querying the graph database. e GEMS soware stack comprises a SPARQL-to-data parallel C++ compiler; a library of distributed data structures; and a custom, multithreaded, runtime system.
