Unordered Associative Containers in STAPL

Abstract

The Standard Template Adaptive Parallel Library (STAPL) is a parallel programming framework for C++ that provides parallel algorithms and containers similar to those found in the Standard Template Library (STL). Currently STAPL is lacking implementations for three unordered associative containers: unordered set, unordered multiset, and unordered multimap. These are commonly used containers in the field of computer science; therefore, their implementations are a necessity for STAPL. The similarity of operations and structure between each container will allow a large portion of code to be reused. The goal of this work is to design and create a parallel implementation of these containers that provides the same user-level facilities as their STL equivalents and displays a high level of scalability when executed on a large number of processors.