USE-RB : Benchmarking how Reasoners Work in Harmony with Modern Hardware

As our computers embed more cores, efficient reasoners are designed with parallelization but also CPU and memory friendliness in mind. These latter contribute to make reasoner tractable in practice despite the computational complexity of logical fragments. However, creating benchmarks to monitor this CPU-friendliness for many reasoners, datasets and logical fragments is a tedious task. In this paper, we present the Universit´e Saint-Etienne Reasoners Benchmark (USE-RB) that automates the setup and execution of reasoners benchmarks with a particular attention to monitor how reasoners work in harmony with the CPU.