Filters








8,135 Hits in 4.7 sec

Remote Memory Access Programming in MPI-3

Torsten Hoefler, James Dinan, Rajeev Thakur, Brian Barrett, Pavan Balaji, William Gropp, Keith Underwood
2015 ACM Transactions on Parallel Computing  
The Message Passing Interface (MPI) 3.0 standard, introduced in September 2012, includes a significant update to the one-sided communication interface, also known as remote memory access (RMA).  ...  In particular, the interface has been extended to better support popular one-sided and global-address-space parallel programming models, to provide better access to hardware performance features, and to  ...  Remote Memory Access Programming in MPI-3 1:5 Memory Exposure MPI RMA offers four calls to expose local memory to remote processes.  ... 
doi:10.1145/2780584 fatcat:5cza22v3irflpfznpaqwfuofqq

Using a cluster as a memory resource: A fast and large virtual memory on MPI

Hiroko Midorikawa, Kazuhiro Saito, Mitsuhisa Sato, Taisuke Boku
2009 2009 IEEE International Conference on Cluster Computing and Workshops  
To access this remote memory, we rely on swap protocols adequate for MPI thread support levels.  ...  This system is an extension of our previously proposed socket-based Distributed Large Memory System (DLM), which offers large virtual memory by using remote memory distributed over nodes in a cluster.  ...  Although the programs shown in Fig. 2 and Fig. 3 appear to be sequential programs, users can implicitly access the remote memory without any knowledge of parallel programming in MPI. III.  ... 
doi:10.1109/clustr.2009.5289180 dblp:conf/cluster/MidorikawaSSB09 fatcat:cfdo2xsfgndn7lzc4e5dinbqei

Enabling highly-scalable remote memory access programming with MPI-3 one sided

Robert Gerstenberger, Maciej Besta, Torsten Hoefler
2013 Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis on - SC '13  
Modern interconnects offer remote direct memory access (RDMA) features. Yet, most applications rely on explicit message passing for communications albeit their unwanted overheads.  ...  The MPI-3.0 standard defines a programming interface for exploiting RDMA networks directly, however, it's scalability and practicability has to be demonstrated in practice.  ...  This work was supported in part by the DOE Office of Science, Advanced Scientific Computing Research, under award number DE-FC02-10ER26011, program manager Lucy Nowell.  ... 
doi:10.1145/2503210.2503286 dblp:conf/sc/GerstenbergerBH13 fatcat:xgxtd45spfdt3ogfwgoe6wo4u4

Enabling Highly-Scalable Remote Memory Access Programming with MPI-3 One Sided

Robert Gerstenberger, Maciej Besta, Torsten Hoefler
2014 Scientific Programming  
Modern interconnects offer remote direct memory access (RDMA) features. Yet, most applications rely on explicit message passing for communications albeit their unwanted overheads.  ...  The MPI-3.0 standard defines a programming interface for exploiting RDMA networks directly, however, it's scalability and practicability has to be demonstrated in practice.  ...  This work was supported in part by the DOE Office of Science, Advanced Scientific Computing Research, under award number DE-FC02-10ER26011, program manager Lucy Nowell.  ... 
doi:10.1155/2014/571902 fatcat:tokihkphivel7m4rheookvir4i

Investigating High Performance RMA Interfaces for the MPI-3 Standard

Vinod Tipparaju, William Gropp, Hubert Ritzdorf, Rajeev Thakur, Jesper L. Traff
2009 2009 International Conference on Parallel Processing  
The MPI-2 Standard, released in 1997, defined an interface for one-sided communication, also known as remote memory access (RMA).  ...  Nonetheless, even 12 years after its existence, the MPI-2 RMA interface remains scarcely used for a number of reasons.  ...  This is not possible in PGAS languages that permit all (or most) of memory to be accessed by remote memory operations.  ... 
doi:10.1109/icpp.2009.54 dblp:conf/icpp/TipparajuGRTT09 fatcat:5ctc5sgk5fgdhdvsojm2bbe7eu

A UPC runtime system based on MPI and POSIX threads

Zhang Zhang, J. Savant, S. Seidel
2006 14th Euromicro International Conference on Parallel, Distributed, and Network-Based Processing (PDP'06)  
Remote memory is accessed by two-sided MPI message passing.  ...  MuPC implements each UPC thread (process) as two Pthreads, one for the user program and private memory accesses, and the other for remote memory accesses.  ...  Acknowledgments The authors wish to thank Brian Wibecan and the UPC group at Hewlett-Packard who gave much assistance in MuPC development.  ... 
doi:10.1109/pdp.2006.16 dblp:conf/pdp/ZhangSS06 fatcat:wnb7onebcvg73nsejis2qpfixq

Problems with using MPI 1.1 and 2.0 as compilation targets for parallel language implementations

Dan Bonachea, Jason Duell
2004 International Journal of High Performance Computing and Networking  
Simulating one-sided communication under the MPI 1.1 API is too expensive, while the MPI 2.0 one-sided API imposes a number of significant restrictions on memory access patterns that that would need to  ...  It would thus seem to offer a convenient "portable network assembly language" to developers of parallel programming languages who wish to target different network architectures.  ...  fraction of all data allocated by GAS programs is typically accessed remotely at some point in program execution).  ... 
doi:10.1504/ijhpcn.2004.007569 fatcat:s6hgbhq6xra7pb2tofekvew27a

Portable, MPI-interoperable coarray fortran

Chaoran Yang, Wesley Bland, John Mellor-Crummey, Pavan Balaji
2014 Proceedings of the 19th ACM SIGPLAN symposium on Principles and practice of parallel programming - PPoPP '14  
5 • MPI-3 Remote Memory Access (RMA) ! 6 Public Private Store Unified MPI_Put MPI_Put Store Separate model Unified model • Does it degrade performance? !  ...  MPI MPI-interoperability • Hard to adopt new programming models in existing applications incrementally! • Interoperable problems in new programming models (examples later)!  ...  CAF-MPI CAF-GASNet ! 20  ... 
doi:10.1145/2555243.2555270 dblp:conf/ppopp/YangBMB14 fatcat:23cyifjo3nc3lpxftnuzp4tdhu

Portable, MPI-interoperable coarray fortran

Chaoran Yang, Wesley Bland, John Mellor-Crummey, Pavan Balaji
2014 SIGPLAN notices  
5 • MPI-3 Remote Memory Access (RMA) ! 6 Public Private Store Unified MPI_Put MPI_Put Store Separate model Unified model • Does it degrade performance? !  ...  MPI MPI-interoperability • Hard to adopt new programming models in existing applications incrementally! • Interoperable problems in new programming models (examples later)!  ...  CAF-MPI CAF-GASNet ! 20  ... 
doi:10.1145/2692916.2555270 fatcat:dd5puu447nanrd3fbwm34o6w6q

Evaluation of Remote Memory Access Communication on the Cray XT3

V. Tipparaju, A. Kot, J. Nieplocha, M. ten Bruggencate, N. Chrisochoides
2007 2007 IEEE International Parallel and Distributed Processing Symposium  
This paper evaluates remote memory access (RMA) communication capabilities and performance on the Cray XT3.  ...  The performance of these interfaces is studied and compared to MPI performance.  ...  The Cray XT3 MPICH2 uses a Portals-based variant of the approach taken in the MPICH2 CH3 ADI3 device to support MPI-2 remote memory access (RMA) [5] .  ... 
doi:10.1109/ipdps.2007.370478 dblp:conf/ipps/TipparajuKNBC07 fatcat:c7cllt26ubb3foqnwbof7paue4

Implementing MPI with the Memory-Based Communication Facilities on the SSS-CORE operating system [chapter]

Kenji Morimoto, Takashi Matsumoto, Kei Hiraki
1998 Lecture Notes in Computer Science  
This paper describes an e cient implementation of MPI on the Memory-Based Communication Facilities; Memory-Based FIFO is used for bu ering by the library, and Remote Write for communication with no bu  ...  They provide low-cost and highly-functional primitives for remote memory accesses. The performance of the library was evaluated on a cluster of workstations connected with a 100Base-TX network.  ...  Summary An MPI library has been implemented for the SSS CORE operating system by the use of the MBCF.  ... 
doi:10.1007/bfb0056579 fatcat:omnugv3k7vhd3mmccujy5k6equ

Design and Implementation of Key Proposed MPI-3 One-Sided Communication Semantics on InfiniBand [chapter]

Sreeram Potluri, Sayantan Sur, Devendar Bureddy, Dhabaleswar K. Panda
2011 Lecture Notes in Computer Science  
As part of MPI-3 effort, the Remote Memory Access group has proposed several extensions to the existing one-sided communication interface to address these limitations.  ...  In this paper, we present design, implementation and evaluation of some of the key one-sided semantics proposed for MPI-3 over InfiniBand, using the MVAPICH2 library.  ...  This research is supported in part by U.S.  ... 
doi:10.1007/978-3-642-24449-0_38 fatcat:gm73opd4uvggbf46fdcjhmvdfe

Evaluating the Impact of Programming Language Features on the Performance of Parallel Applications on Cluster Architectures [chapter]

Konstantin Berlin, Jun Huan, Mary Jacob, Garima Kochhar, Jan Prins, Bill Pugh, P. Sadayappan, Jaime Spacco, Chau-Wen Tseng
2004 Lecture Notes in Computer Science  
We compare a number of programming languages (Pthreads, OpenMP, MPI, UPC, Global Arrays) on both shared and distributed-memory architectures.  ...  Powerful compiler analysis and optimization can help reduce software overhead, but features such as fine-grain remote accesses are inherently expensive on clusters.  ...  Conclusions In this paper, we examined language features from a number of parallel programming paradigm/languages (MPI, UPC, OpenMP, Java, C/Pthreads, Global Arrays) for their performance and ease of use  ... 
doi:10.1007/978-3-540-24644-2_13 fatcat:js24djykkfhohk2gmc2m4dmbdu

Notified Access: Extending Remote Memory Access Programming Models for Producer-Consumer Synchronization

Roberto Belli, Torsten Hoefler
2015 2015 IEEE International Parallel and Distributed Processing Symposium  
Remote Memory Access (RMA) programming enables direct access to low-level hardware features to achieve high performance for distributed-memory programs.  ...  We implement our scheme in an open source MPI-3 RMA library and demonstrate lower overheads (two cache misses) than other point-to-point synchronization mechanisms.  ...  Kathy Yelick (LBNL), Edgar Solomonik, Timo Schneider, and Salvatore Di Girolamo for helpful discussions, Larry Kaplan (Cray) for help with uGNI, and the Swiss National Supercomputing Centre (CSCS) for access  ... 
doi:10.1109/ipdps.2015.30 dblp:conf/ipps/BelliH15 fatcat:vihdr3456zd5popdbgdd5h7hdi

MC-Checker: Detecting Memory Consistency Errors in MPI One-Sided Applications

Zhezhe Chen, James Dinan, Zhen Tang, Pavan Balaji, Hua Zhong, Jun Wei, Tao Huang, Feng Qin
2014 SC14: International Conference for High Performance Computing, Networking, Storage and Analysis  
This paper presents MC-Checker, a new tool that detects memory consistency errors in MPI one-sided applications.  ...  While such interfaces can be extremely efficient, they also impose challenges in properly performing asynchronous accesses to shared data.  ...  This work was supported in part by the U.S. NSF grants #CCF-0953759 (CAREER Award), #CCF-1218358, and #CCF-1319705, by the U.S.  ... 
doi:10.1109/sc.2014.46 dblp:conf/sc/ChenDTBZWHQ14 fatcat:fa2ooryt2zc5dohmkagfrm2wom
« Previous Showing results 1 — 15 out of 8,135 results