LUBM (Lehigh University Benchmark) was one of the earlies standardised SPARQL benchmarks, and while no longer as relevant to modern SPARQL engines, drove a lot of early optimisation research and development.
This version of the data generator was a rewrite of the original single-threaded data generator designed to generate identical output in parallel. This allows generating datasets for very large benchmarks much more rapidly than was possible with the original data generator.
This tool was developed while working for YarcData™ (a Cray Inc. company) and was open sourced in July 2015.
Source Code for the benchmarker may be obtained from the GitHub repository.