A Parallel Task-based Approach to Linear Algebra

TL;DR

This paper introduces a new parallel task-based model using GPRM for linear algebra problems, demonstrating improved performance and flexibility over OpenMP on many-core systems.

Contribution

It proposes an alternative task management model based on GPRM and applies it to LU factorization, showing advantages over OpenMP in efficiency and stability.

Findings

Performance improvement over OpenMP in LU factorization

Enhanced task management efficiency and stability

Successful deployment on TILEPro64 system

Abstract

Processors with large numbers of cores are becoming commonplace. In order to take advantage of the available resources in these systems, the programming paradigm has to move towards increased parallelism. However, increasing the level of concurrency in the program does not necessarily lead to better performance. Parallel programming models have to provide flexible ways of defining parallel tasks and at the same time, efficiently managing the created tasks. OpenMP is a widely accepted programming model for shared-memory architectures. In this paper we highlight some of the drawbacks in the OpenMP tasking approach, and propose an alternative model based on the Glasgow Parallel Reduction Machine (GPRM) programming framework. As the main focus of this study, we deploy our model to solve a fundamental linear algebra problem, LU factorisation of sparse matrices. We have used the SparseLU benchmark from the BOTS benchmark suite, and compared the results obtained from our model to those of the OpenMP tasking approach. The TILEPro64 system has been used to run the experiments. The results are very promising, not only because of the performance improvement for this particular problem, but also because they verify the task management efficiency, stability, and flexibility of our model, which can be applied to solve problems in future many-core systems.