Survey of MPI Implementations

Hafeez, Mehnaz; Asghar, S.; Malik, Usman Ahmad; Rehman, Adeel Ur; Riaz, Naveed

doi:10.1007/978-3-642-22027-2_18

Cited by 7 publications

(3 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To provide the experimental work (in particular, the application model) and the proof of concepts, we have used our prototype tool [12], which is implemented in C++ and targets MPI applications based on OpenMPI [26]. This prototype tool models and analyses the whole application execution using MRNet, which provides efficient multicast and reduction communications for parallel and distributed tools [27].…”

Section: Experimental Workmentioning

confidence: 99%

Online root-cause performance analysis of parallel applications

2015

View full text Add to dashboard Cite

Section: Experimental Workmentioning

confidence: 99%

Online root-cause performance analysis of parallel applications

2015

View full text Add to dashboard Cite

“…The tool is implemented in C++ and targets MPI applications. We built our prototype tool using OpenMPI [15], Dyninst API [47] and MRNet [46]. It has been tested on x86/Linux, Intel IA64/Linux, and partially on PowerPC-64/Linux.…”

Section: Prototype For Dynamic Abstraction Of Mpi Applicationsmentioning

confidence: 99%

Automated and dynamic abstraction of MPI application performance

2016

View full text Add to dashboard Cite

Developing an efficient parallel application is not an easy task, and achieving a good performance requires a thorough understanding of the program's behavior. Careful performance analysis and optimization are crucial. To help developers or users of these applications to analyze the program's behavior, it is necessary to provide them with an abstraction of the application performance. In this paper, we propose a dynamic performance abstraction technique, which enables the automated discovery of causal execution paths, composed of communication and computational activities, in MPI parallel programs. This approach enables autonomous and low-overhead execution monitoring that generates performance knowledge about application behavior for the purpose of online performance diagnosis. Our performance abstraction technique reflects an application behavior and is made up of elements correlated with highlevel program structures, such as loops and communication operations. Moreover, it characterizes all elements with statistical execution profiles. We have evaluated our approach on a variety of scientific parallel applications. In all scenarios, our online performance abstraction technique proved effective for low-overhead capturing of the program's behavior and facilitated performance understanding.

show abstract

“…There is not enough work carried out for secured MPI, though, there are many implementations of MPI and MPI-2 available [18]. The general approach to improve security of MPI applications may allow application programmers themselves to include message confidentiality that reduces portability and flexibility.…”

Section: Literature Reviewmentioning

confidence: 99%

Secure Peer to Peer Message Passing using A-JUMP

Asghar¹,

Hafeez²,

Malik³

et al. 2011

Proceedings of the International Symposium on Grids and Clouds and the Open Grid Forum — PoS(ISGC 2011 &Amp; OGF 31)

Self Cite

View full text Add to dashboard Cite

MPI is a de facto standard for message passing for high performance parallel, as well as, for distributed computing environment. The static and homogenous model of MPI is not compatible with the dynamic and heterogeneous Grid environment. There are not many implementations which offer message passing over Internet and Grids. P2P-MPI and A-JUMP are MPI implementations, which provide both point-to-point and collective data operations over Internet using Java. However, none of these MPI implementations are categorized as a secure message passing implementation. In these MPI implementations security is overlooked to avoid performance degradation. Security becomes a fundamental concern when considering message passing over Internet, and should be accordingly addressed. Moreover, access over Internet requires authorization/authentication of resources and users along with encryption of data. In this paper, we have proposed secure message passing over Internet and Grids using A-JUMP. This paper also presents performance analysis and comparison of the proposed model with P2P-MPI and A-JUMP.

show abstract

Survey of MPI Implementations

Cited by 7 publications

References 27 publications

Online root-cause performance analysis of parallel applications

Online root-cause performance analysis of parallel applications

Automated and dynamic abstraction of MPI application performance

Secure Peer to Peer Message Passing using A-JUMP

Contact Info

Product

Resources

About