Hybrid MPI/OpenMP Parallel Programming on Clusters of Multi-Core SMP Nodes

Rabenseifner, Rolf; Hager, Georg; Jost, Gabriele

doi:10.1109/pdp.2009.43

Cited by 304 publications

(188 citation statements)

References 5 publications

Supporting

Mentioning

182

Contrasting

Unclassified

Order By: Relevance

“…The last 3-4 years however, another strong alternative has evolved; the MPI 3.0 Shared Memory support mechanism, which improves significantly the previous existed Remote Memory Access utilities of MPI, towards the direction of optimized operation inside a multicore node. As analyzed in [9][10][11] both the above referred hybrid models (MPI+OpenMP, MPI+MPI 3.0 Shared Memory) have their pros and cons, and it's not straightforward that they outperform pure MPI implementations in all cases. Among all the alternatives the MPI+OpenMP hybrid approach is still regarded as the most efficient one, however the MPI+MPI 3.0 Shared Memory approach is highly competitive.…”

Section: Introductionmentioning

confidence: 99%

Simplex Parallelization in a Fully Hybrid Hardware Platform

Mamalis¹,

Perlitis²

2017

ijacsa

View full text Add to dashboard Cite

Abstract-The simplex method has been successfully used in solving linear programming (LP) problems for many years. Parallel approaches have also extensively been studied due to the intensive computations required, especially for the solution of large LP problems. Furthermore, the rapid proliferation of multicore CPU architectures as well as the computational power provided by the massive parallelism of modern GPUs have turned CPU / GPU collaboration models increasingly into focus over the last years for better performance. In this paper, a highly scalable implementation framework of the standard full tableau simplex method is first presented, over a hybrid parallel platform which consists of multiple multicore nodes interconnected via a high-speed communication network. The proposed approach is based on the combined use of MPI and OpenMP, adopting a suitable column-based distribution scheme for the simplex tableau. The parallelization framework is then extended in such a way that it can exploit concurrently the full power of the provided resources on a multicore single-node environment with a CUDA-enabled GPU (i.e. using the CPU cores and the GPU concurrently), based on a suitable hybrid multithreading/GPU offloading scheme with OpenMP and CUDA. The corresponding experimental results show that the hybrid MPI+OpenMP based parallelization scheme leads to particularly high speed-up and efficiency values, considerably better than in other competitive approaches, and scaling well even for very large / huge linear problems. Furthermore, the performance of the hybrid multithreading/GPU offloading scheme is clearly superior to both the OpenMP-only and the GPU-only based implementations in almost all cases, which validates the worth of using both resources concurrently. The most important, when it is used in combination with MPI in a multi-node (fully hybrid) environment, it leads to substantial improvements in the speedup achieved for large and very large LP problems.

show abstract

Section: Introductionmentioning

confidence: 99%

Simplex Parallelization in a Fully Hybrid Hardware Platform

Mamalis¹,

Perlitis²

2017

ijacsa

View full text Add to dashboard Cite

show abstract

“…Such a topology discovery feature would be very useful for many other applications, especially MPI process managers. Hybrid MPI+OpenMP applications could also take advantage from it [15]. There is a clear need for a portable and accurate tool dedicated to topology discovery.…”

Section: Related Work and Discussionmentioning

confidence: 99%

Towards an Efficient Process Placement Policy for MPI Applications in Multicore Environments

Mercier

Clet-Ortega

2009

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. This paper presents a method to efficiently place MPI processes on multicore machines. Since MPI implementations often feature efficient supports for both shared-memory and network communication, an adequate placement policy is a crucial step to improve applications performance. As a case study, we show the results obtained for several NAS computing kernels and explain how the policy influences overall performance. In particular, we found out that a policy merely increasing the intranode communication ratio is not enough and that cache utilization is also an influential factor. A more sophisticated policy (eg. one taking into account the architecture's memory structure) is required to observe performance improvements.

show abstract

“…OpenMPparadigm is designed for shared-memory, MPIis adopt within or across nodes, and MPI+OpenMP hybrid paradigm ( [26]) is employed for exploiting multilevel of parallelism. Recently, great efforts have been laid on the incorporation of MPI-enabled paradigm with remote sensing data processing in the large scale scenarios.…”

Section: Cluster Computing For Rs Data Processingmentioning

confidence: 99%

pipsCloud: High performance cloud computing for remote sensing big data management and processing

Wang

Yan

et al. 2018

Future Generation Computer Systems

173

View full text Add to dashboard Cite

With the increasing requirement of accurate and up-to-date resource & environmental information for regional and global monitoring, large-region covered multi-temporal, multi-spectral massive remote sensing (RS) datasets are exploited for processing. The remote sensing data processing generally follows a complex multi-stage processing chain, which consists of several independent processing steps subject to types of RS applications. In general the RS data processing for regional environmental and disaster monitoring are recognized as typical both compute-intensive and data-intensive applications.To solve the aforementioned issues efficiently, we propose pipsCloud which combine recent Cloud computing and HPC techniques to enable large-scale RS data processing system as on-demand real-time services. Benefiting from the ubiquity, elasticity and high-level of transparency of Cloud computing model, the massive RS data managing and data processing for dynamic environmental monitoring are all encapsulate as Cloud with Web interfaces. Where, a Hilbert-R + based data indexing mechanism is employed for optimal query and access of RS imageries, RS data products as well as interim data. In the core platform beneath the Cloud services, we provide a parallel file system for massive high-dimensional RS data and offers interfaces for intensive irregular RS data accessing so as to provide improved data locality and optimized I/O performance. Moreover, we adopt an adaptive RS data analysis workflow manage system for on-demand workflow construction and collaborative execution of distributed complex chain of RS data processing, such as forest fire detection, mineral resources and coastline monitoring. Through the experimental analysis we have show the efficiency of the pipsCloud platform.

show abstract

Hybrid MPI/OpenMP Parallel Programming on Clusters of Multi-Core SMP Nodes

Cited by 304 publications

References 5 publications

Simplex Parallelization in a Fully Hybrid Hardware Platform

Simplex Parallelization in a Fully Hybrid Hardware Platform

Towards an Efficient Process Placement Policy for MPI Applications in Multicore Environments

pipsCloud: High performance cloud computing for remote sensing big data management and processing

Contact Info

Product

Resources

About