CHiPPS: A Framework for Implementing Parallel Search Algorithms

The COIN-OR High Performance Parallel Search (CHiPPS) framework is a C++ class library for implementing and parallelizing tree search algorithms. It currently consists of a library of base classes known as the Abstract Library for Parallel Search (ALPS) and two other layers useful for implementing algorithms based on mathematical programming paradigms. Because of its general approach, however, ALPS supports the implementation of a wide variety of tree search algorithms and applications by creating derived classes implementing specific algorithmic components. CHiPPS employs a number of features to improve scalability and is designed specifically to support the implementation of data intensive algorithms, in which large amounts of knowledge are generated and must be maintained and shared during the search. Implementing such algorithms in a scalable manner is challenging both because of storage requirements and because of communications overhead incurred in the sharing of data. In this abstract, we describe the design of CHiPPS, focusing primarily on the ALPS layer, and indicate how the design addresses these challenges.