Filters








1,502 Hits in 5.0 sec

Irregular Redistribution Scheduling by Partitioning Messages [chapter]

Chang Wu Yu, Ching-Hsien Hsu, Kun-Ming Yu, C. -K. Liang, Chun-I Chen
2005 Lecture Notes in Computer Science  
This work presents an efficient algorithm to partition large messages into multiple small ones and schedules them by using the minimum number of steps without communication contention and, in doing so,  ...  In irregular redistribution, however, messages with varying sizes are transmitted in the same communication step.  ...  In irregular redistribution, messages of varying sizes are scheduled in the same communication step.  ... 
doi:10.1007/11572961_24 fatcat:2lpuebmb2vd7zgkcpns2sstaje

The Paradigm compiler for distributed-memory multicomputers

P. Banerjee, J.A. Chandy, M. Gupta, E.W. Hodges, J.G. Holm, A. Lain, D.J. Palermo, S. Ramaswamy, E. Su
1995 Computer  
A unified approach efficiently supports regular and irregular computations using data and functional parallelism.  ...  parallel distributed-memory multicomputers can achieve the high performance levels required to solve the Grand Challenge computational science problems (a class of computational applications, identified by  ...  .-~~ This research was supported in part by the Office of Naval Research under Contract N00014-91J-1096, the National Aeronautics and Space Administration under Contract NASA NAG l-613, an AT&T graduate  ... 
doi:10.1109/2.467577 fatcat:ghmtervcfzehzlelvf2ealwgyu

Optimizing Scheduling Stability for Runtime Data Alignment [chapter]

Ching-Hsien Hsu, Chao-Yang Lan, Shih-Chang Chen
2006 Lecture Notes in Computer Science  
The proposed technique introduces a new scheduling policy, Local Message Reduction (LMR), to obtain better communication schedule adaptive to different environments. o evaluate the performance of the proposed  ...  In this paper, we present an algorithm independent technique for optimizing scheduling stability of different scheduling heuristics.  ...  Phase 2: Identify all local messages in a given irregular data redistribution instance. Phase 3: Reduce the communication cost of those messages found in last phase by the factor RLR.  ... 
doi:10.1007/11807964_83 fatcat:xbehlj5l6naezda4lqos4zkey4

Runtime and language support for compiling adaptive irregular programs on distributed-memory machines

Yuan-Shin Hwang, Bongki Moon, Shamik D. Sharma, Ravi Ponnusamy, Raja Das, Joel H. Saltz
1995 Software, Practice & Experience  
Besides providing runtime support to users, CHAOS can also be used by compilers to automatically parallelize irregular applications.  ...  This paper presents a library called CHAOS, which helps users implement irregular programs on distributed-memory message-passing machines, such as the Paragon, Delta, CM-5 and SP-1.  ...  Fox, Alok Choudhary, and Sanjay Ranka for many enlightening discussions, and Chuck Koelbel, Ken Kennedy and Seema Hiranandani for many useful discussions about integrating Fortran D runtime support for irregular  ... 
doi:10.1002/spe.4380250603 fatcat:zbzluzrtt5futjpv4m3bbumdjy

A Divide-and-Conquer Algorithm for Irregular Redistribution in Parallelizing Compilers

Hui Wang, Minyi Guo, Daming Wei
2004 Journal of Supercomputing  
steps for irregular array redistribution.  ...  This paper is devoted to develop an efficient algorithm that attempts to obtain near optimal scheduling while satisfying the conditions of minimal message size of total steps and the minimal number of  ...  the minimal number of steps for irregular array redistribution.  ... 
doi:10.1023/b:supe.0000026848.34971.27 fatcat:mvqezosspve5bfyzc7g6skajle

Runtime Support for Parallelization of Data-Parallel Applications on Adaptive and Nonuniform Computational Environments

Maher Kaddoura, Sanjay Ranka
1997 Journal of Parallel and Distributed Computing  
The approach presented is reasonably general and is applicable to a wide variety of regular as well as irregular applications.  ...  The number of messages generated can also be taken into account by incorporating it into the cost of redistribution.  ...  The number of messages generated by our library could be reduced signi cantly by using multicast.  ... 
doi:10.1006/jpdc.1997.1340 fatcat:7on3dqufyfhlxhibi5wnzrpun4

Runtime support for parallelization of data-parallel applications on adaptive and nonuniform computational environments

M. Kaddoura, S. Ranka
1996 Proceedings of 5th IEEE International Symposium on High Performance Distributed Computing HPDC-96  
The approach presented is reasonably general and is applicable to a wide variety of regular as well as irregular applications.  ...  The number of messages generated can also be taken into account by incorporating it into the cost of redistribution.  ...  The number of messages generated by our library could be reduced signi cantly by using multicast.  ... 
doi:10.1109/hpdc.1996.546171 dblp:conf/hpdc/KaddouraR96 fatcat:omwle7f4sjcsloa2mbbw4cgm5y

New Bipartite Graph Techniques for Irregular Data Redistribution Scheduling

Li, Yu
2019 Algorithms  
In general, an array can be distributed to multiple processing systems by using regular or irregular distributions.  ...  For many parallel and distributed systems, automatic data redistribution improves its locality and increases system performance for various computer problems and applications.  ...  In [15] , Yu et al. presents an efficient algorithm to partition large messages into multiple small ones and schedules them by using the minimum number of steps without communication contention and, in  ... 
doi:10.3390/a12070142 fatcat:lfpxyud5nbgarp4xxe4oxulzbu

Run-time and compile-time support for adaptive irregular problems

Shamik D. Sharma, Ravi Ponnusamy, Bongki Moon, Yuan Shin Hwang, Raja Das, Joel Saltz
1994 Supercomputing, Proceedings  
Implementingsuch problems on distributed memory machines requires support for dynamic data partitioning, e cient preprocessing and fast data migration.  ...  It subsumes the previous PARTI library which targeted only static irregular problems.  ...  applicable partitioners and how to embed such partitioners in compilers; also Chuck Koelbel, Ken Kennedy and Seema Hiranandani for many useful discussions about integrating Fortran-D runtime support for irregular  ... 
doi:10.1145/602791.602793 fatcat:egnsgqwi4rdshfg7li367ks5e4

Dynamic Data Distribution and Processor Repartitioning for Irregularly Structured Computations

José E. Moreira, Vijay K. Naik, Samuel P. Midkiff
1998 Journal of Parallel and Distributed Computing  
We also present performance levels achieved by the data redistribution primitives, using a sparse Cholesky factorization algorithms as a model irregular application.  ...  In this paper, we discuss DRMS support for developing reconfigurable irregular applications and describe the dynamic data redistribution mechanisms in some detail.  ...  At an SOP, the application partition can be reconfigured either by an internal application request or by an action from the system resource scheduler.  ... 
doi:10.1006/jpdc.1998.1436 fatcat:nqkflvrknjctbhlpp7bw6tmive

SUPPLE: An efficient run-time support for non-uniform parallel loops

Salvatore Orlando, Raffaele Perego
1999 Journal of systems architecture  
haracterized by poor performance.  ...  The idea behind this feature of the CHAOS library is the run{time redistribution of arrays, and the consequent re{mapping of iterations.  ...  Another problem considered in SUPPLE is data coherence when hybrid scheduling is adopted: in fact some tiles of a BLOCK partition may beupdated by processors other than the owner of the partition itself  ... 
doi:10.1016/s1383-7621(98)00071-x fatcat:kiu4ma2g5bf2na72mnmf6xsatq

A Comparison of Implementation Strategies for Nonuniform Data-Parallel Computations

Salvatore Orlando, Raffaele Perego
1998 Journal of Parallel and Distributed Computing  
To reduce overheads, the compiler partitions the computations among the processors at compile{time, on the basis of the static data distribution suggested by the programmer.  ...  Workload imbalance can be mitigated by cyclically distributing data and associated computations, or by employing adaptive strategies which build a more balanced schedule at run{time, on the basis of the  ...  Overhead minimization is accomplished by avoiding sending several messages or replicated data to the same processor, i.e. by exploiting message vectorization, coalescing, and aggregation optimizations.  ... 
doi:10.1006/jpdc.1998.1456 fatcat:dpu3rbatdned7evd5pvvmmnsii

VFC: The Vienna Fortran Compiler

Siegfried Benkner
1999 Scientific Programming  
This comprises data locality assertions, non‐local access specifications and the possibility of reusing runtime‐generated communication schedules of irregular loops.  ...  achieved even for highly irregular codes.  ...  Moreover, VFC supports dynamic data redistribution either by means of the REDISTRIBUTE directive or by redistribution as a result of procedure calls.  ... 
doi:10.1155/1999/304639 fatcat:rqcpi2ih4veynjqqxwxnrqg57a

Supporting irregular distributions using data-parallel languages

R. Ponnusamy, Yuan-Shin Hwang, R. Das, J.H. Saltz, A. Choudhary, G. Fox
1995 IEEE Parallel & Distributed Technology Systems & Applications  
These partitioned data arrays are called distributed arrays. Many applications can be efficiently implemented by using simple schemes for mapping distributed arrays.  ...  g Languages such as Fortran D provide irregular distribution schemes that can efficiently support irregular-problems. Irregular distributions can also be emulated in HPF.  ...  applicability of HPF intrinsics and extrinsics fo;irre&i prodlems; we also thank Ken Kennedy, Seetna Hiranandani, and Sanjay Ranka for many useful discussions about intkgrating Forum D runtime support for irregular  ... 
doi:10.1109/88.384581 fatcat:qscb4olexveehlp7swtnbm3ymm

Distributed Parallel Inference on Large Factor Graphs [article]

Joseph E. Gonzalez, Yucheng Low, Carlos E. Guestrin, David O'Hallaron
2012 arXiv   pre-print
We develop a new efficient parallel inference algorithm, DBRSplash, which incorporates over-segmented graph partitioning, belief residual scheduling, and uniform work Splash operations.  ...  and Joseph Gonzalez is supported by the AT&T Labs Fellowship.  ...  Acknowledgements This work is supported by ONR Young Investigator Program grant N00014-08-1-0752, the ARO under MURI W911NF0810242, DARPA IPTO FA8750-09-1-0141, the NSF under grants NeTS-NOSS and CNS-0625518  ... 
arXiv:1205.2645v1 fatcat:hpbcbyv3i5gt5iflrjnyirxwum
« Previous Showing results 1 — 15 out of 1,502 results