Skip to main content

OpenMP – Moving Beyond Shared-Memory Parallelism

Speaker: Henry Jin, Computational Technologies Branch, NASA Ames Research Center

Presentation

Seminar Slide Deck (PDF-3.2MB)

Abstract

OpenMP has been the de-facto standard for shared-memory parallel programming since its first release in 1997. As the current trend of parallel systems is towards a cluster of multi-core nodes enhanced with accelerators, software development for such systems has become a major challenge. Both low-level and high-level programming models have been developed to address complex hierarchical structures at different hardware levels and to ease the programming effort. In this talk, we will present the recent development in OpenMP 4.0 for supporting accelerated devices and vector parallelism. We will also discuss some of the challenges in achieving desired performance as well as future effort in the language standard.

Biography

Henry Jin obtained his Ph.D. in Physics from the University of Tennessee in 1991. He currently is a senior member of the Application Performance and Productivity Group in NASA Advanced Supercomputing (NAS) Division. Since initially joined the Parallel Tools Group at NASA Ames in 1996, he has been working on many projects related to parallel computing for distributed and shared memory systems, and authored and/or co-authored more than 60 peer-reviewed publications in the relevant areas. His expertise includes in-depth knowledge of parallel programming languages and libraries, design of parallel benchmarks, and the development of parallelization and optimization tools for scientific applications. He is an internationally recognized expert in the OpenMP programming model and has served as a NASA representative on the OpenMP Architecture Review Board since 2005. His research interests include parallel programming methodology for homogeneous and heterogeneous systems, software tools for performance optimization of parallel applications, and benchmarking parallel systems.