“…To tackle this predicament, researchers need solutions that are of a fast, resource‐efficient, and scalable nature (Pfeifer et al, 2014; Rozas et al, 2017; Szpiech & Hernandez, 2014). However, popular computational tools, though capable of being executed on HPC (high‐performance computing) clusters, have been designed with the goal of processing data that spans a few hundred individuals across thousands of polymorphic sites (Cook & Andersen, 2017; Ko et al, 2020; Siepel, 2019). They use single‐thread algorithms that do not exploit the prevalent parallel processing technologies made available through multi‐core CPUs (central processing unit), and GPUs (graphical processing unit) (Cook & Andersen, 2017; Ghorpade, 2012; Pfeifer et al, 2014; Tendler et al, 2002).…”