Evolution of a minimal parallel programming model. (January 2018)

• Record Type: Journal Article
• Authors: Lusk, Ewing; Butler, Ralph; Pieper, Steven C
• Other Names: Dongarra Jack guest-editor.; Tourancheau Bernard guest-editor.
• Is Part Of: International journal of high performance computing applications. Volume 32:Number 1(2018)
• Page Start: 4
• Page End: 13
• Publication Date: 2018-01
• Subjects: Parallel computing -- programming models -- load balancing -- automated theorem proving -- nuclear physics; High performance computing -- Periodicals; Supercomputers -- Periodicals; 004.1105
• DOI: 10.1177/1094342017703448
• Languages: English
• ISSNs: 1094-3420

Abstract

We take a historical approach to our presentation of self-scheduled task parallelism, a programming model with its origins in early irregular and nondeterministic computations encountered in automated theorem proving and logic programming. We show how an extremely simple task model has evolved into a system, asynchronous dynamic load balancing (ADLB), and a scalable implementation capable of supporting sophisticated applications on today's (and tomorrow's) largest supercomputers; and we illustrate the use of ADLB with a Green's function Monte Carlo application, a modern, mature nuclear physics code in production use. Our lesson is that by surrendering a certain amount of generality and thus applicability, a minimal programming model (in terms of its basic concepts and the size of its application programmer interface) can achieve extreme scalability without introducing complexity.