Senior Principal Software Engineer in the Oracle Linux and Virtualization engineering organization at Oracle Corporation

“What Could Possibly Go Wrong Using OpenMP?”

Tuesday 14 February, 1.30pm – 2.05pm

Abstract

OpenMP has established itself as the standard for shared memory parallel programming in HPC. One of the main advantages is the ease of use. A simple pragma added to the source code is sufficient to parallelize a block of code. All the complexity behind this, is hidden in the compiler and run time library.

This ease of use also has a somewhat darker side though. As easy as it is to parallelize code, it is also just as easy to make a mistake that affects the correctness of results, or the performance.

In this talk, we start with a very quick overview of OpenMP and explain the concepts needed in the remainder of this talk. This is followed by some common pitfalls and the solution. We end with a case study that shows how a targeted approach and tuning strategy led to a dramatic improvement in the performance of an application from graph analysis.

Bio

Ruud is a Senior Principal Software Engineer in the Oracle Linux and Virtualization engineering organization at Oracle Corporation.

He operates on a world-wide basis. His key area of expertise is in parallel computing, in particular on shared memory systems using OpenMP. He has been involved with OpenMP from the start and has extensive experience in highly scalable shared memory performance on various types of algorithms, including graph analysis.

He also regularly gives technical presentations and tutorials at conferences and workshops.

Ruud has studied mathematics and physics and has been involved with the performance of technical applications for 25+ years. Before joining Oracle he worked at Sun Microsystems, SGI, Convex Computer Corporation, the University of Utrecht and Phillips.

He is also a board member of IDC’s HPC Advisory Committee, regularly receives invitations to be on the program committee of international conferences and has a strong interest in Interval Analysis and Interval Arithmetic.

Ruud’s current area of interest is the gprofng application profiling tool that is part of the GNU binutils tool suite. In addition to his code development work, he is actively involved with the promotion of this tool, including giving presentations.

Ruud has published over 20 conference papers related to application tuning and parallelization, several technical white papers and is co-author of the books “Using OpenMP”, and “Using OpenMP – The Next Step”, both published by MIT Press.

Slides