Tutorial on Hybrid MPI and OpenMP Parallel Programming
Level: 25% Introductory, 50% Intermediate, 25% Advanced
Rolf Rabenseifner, HLRS, Germany
Georg Hager, University of Erlangen-Nuremberg, Germany
Gabriele Jost, Sun Microsystems, Germany
Rainer Keller, HLRS, Germany
Abstract. Most HPC systems are clusters of shared memory nodes. Such systems can be PC clusters with dual or quad boards, but also “constelation” type systems with large SMP nodes. Parallel programming must combine the distributed memory parallelization on the node inter-connect with the shared memory parallelization inside of each node.
This tutorial analyzes the strength and weakness of several parallel programming models on clusters of SMP nodes. Various hybrid MPI+OpenMP programming models are compared with pure MPI. Benchmark results of several platforms are presented. A hybrid-masteronly programming model can be used more efficiently on some
vector-type systems, but also on clusters of dual-CPUs. On other systems, one CPU is not able to saturate the inter-node network and the commonly used masteronly programming model suffers from insufficient inter-node bandwidth. The thread-safety quality of several existing MPI libraries is also discussed. Case studies from the fields of CFD (NAS Parallel Benchmarks and Multi-zone NAS Parallel Benchmarks, in detail), Climate Modelling (POP2, maybe) and Particle Simulation (GTC, maybe) will be provided to demonstrate various aspect of hybrid MPI/OpenMP programming.
Another option is the use of distributed virtual shared-memory technologies which enable the utilization of “near-standard” OpenMP on distributed memory architectures. The performance issues of this approach and its impact on existing applications are discussed. This tutorial analyzes strategies to overcome typical drawbacks of
easily usable programming schemes on clusters of SMP nodes.
About the speakers.
Rolf Rabenseifner studied mathematics and physics at the University of Stuttgart. Since 1984, he has worked at the High-Performance Computing-Center Stuttgart (HLRS). He led the projects DFN-RPC, a remote procedure call tool, and MPI-GLUE, the first metacomputing MPI combining different vendor’s MPIs without loosing the full MPI interface. In his dissertation, he developed a controlled logical clock as global time for trace-based profiling of parallel and distributed applications. Since 1996, he has been a member of the MPI-2 Forum. From January to April 1999, he was an invited researcher at the Center for High-Performance Computing at Dresden University of Technology. Currently, he is head of Parallel Computing – Training and Application Services at HLRS. He is involved in MPI profiling and benchmarking, e.g., in the HPC Challenge Benchmark Suite. In recemt projects, he studied parallel I/O, parallel programming models for clusters of SMP nodes, and optimization of MPI collective routines. In workshops and summer schools, he teaches parallel programming models in many universities and labs in Germany
Georg Hager studied theoretical physics at the University of Bayreuth, specializing in nonlinear dynamics. Since 2000 he is a member of the HPC Services group at the Regional Computing Center Erlangen (RRZE), which is part of the University of Erlangen-Nuernberg. His daily work encompasses all aspects of user support in High Performance Computing like tutorials and training, code parallelization, profiling and optimization and the assessment of novel computer architectures and tools. In his dissertation he developed a shared-memory parallel density-matrix renormalization group algorithm for ground-state calculations in strongly correlated electron systems. Recent work includes architecture-specific optimization strategies for current microprocessors and special topics in shared memory programming.Gabriele Jost obtained her doctorate in Applied Mathematics from the University of Goettingen, Germany. For more than a decade she worked for various vendors (Suprenum GmbH, Thinking Machines Corporation, and NEC) of high performance parallel computers in the areas of vectorization, parallelization, performance analysis and optimization of scientific and engineering applications. In 1998 she joined the NASA Ames Research Center in Moffett Field, Califorina, USA as a Research Scientist. Here her work focused on evaluating and enhancing tools for parallel program development and investigating the usefulness of differentparallel programming paradigms. In 2005 she moved from Califorina to the Pacific Northwest and joined Sun Microsystems as a staff engineer in the Compiler Performance Engineering team. Her task is the analysis of compiler generated code and providing feedback and suggestions for improvement to the compiler group. Her research interest remains in area of performance analysis and evaluation of programming paradigms for high performance computing.
Rainer Keller is a scientific employee at the High-Performance Computing Centre Stuttgart (HLRS) since 2001. He earned his diploma in Computer Science at the University of Stuttgart. Currently, he is the head of the group Applications, Models and Tools at the HLRS. His professional interest are Parallel Computation using and working on MPI with Open MPI and shared memory parallelisation with OpenMP, as well as distributed computing using the MetaComputing Library PACX-MPI. His work includes performance analysis and optimization of parallel applications, as well as the assessment of and porting to new hardware technologies, including the training of HLRS users in parallel application development. He is involved in several European projects, such as HPC-Europa.