Group Communication Patterns for High Performance Computing in Scala

ABSTRACT. We developed a novel Functional Object-Oriented PARallel framework (FOOPAR) for analyzable high-level high-performance computing in Scala. Central to this framework are Distributed Memory Parallel Data structures, i.e., collections of data distributed in a shared nothing system together with parallel operations on these data.

In this position paper we shortly present FOOPAR’s design and focus on its properties w.r.t. analyzability and correctness. We prove the correctness and safety of one communication algorithm and show how specification testing (via ScalaCheck) can be used to bridge the gap between proof and implementation.