Execution time support for adaptive scientific algorithms on distributed memory machines

Berryman, Harry; Saltz, Joel; Scroggs, Jeffrey S.

doi:10.1002/cpe.4330030303

Cited by 55 publications

(17 citation statements)

References 9 publications

(9 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…4 The 4 We use a version of Em3d that uses arrays for storing electric and magnetic field values and is written in a style that is more akin to Fortran and HPF programming models. Hence, the program has performance characteristics that are different from those of the pointer-based versions described in [8].…”

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Analyses and Optimizations for Shared Address Space Programs

Krishnamurthy

Yelick

1996

Journal of Parallel and Distributed Computing

View full text Add to dashboard Cite

Section: Resultsmentioning

confidence: 99%

“…In Fig. 7 4 ], implies that a 1 precedes a 6 for any execution of the program. Since a simple-path to a 1 corresponds to a runtime execution where all the accesses in the sequence execute before a 1 , a 6 will never occur in a simple-path to a 1 .…”

Section: Analyzing Barrier Synchronizationmentioning

confidence: 98%

Analyses and Optimizations for Shared Address Space Programs

Krishnamurthy

Yelick

1996

Journal of Parallel and Distributed Computing

View full text Add to dashboard Cite

“…The PARTI/CHAOS libraries [6], [31] facilitated manual development of parallel message-passing code. Compiler support for optimizing communications within the executor was also explored [1], [44].…”

Section: Related Workmentioning

confidence: 99%

Code generation for parallel execution of a class of irregular loops on distributed memory systems

Ravishankar¹,

Eisenlohr²,

Pouchet³

et al. 2012

2012 International Conference for High Performance Computing, Networking, Storage and Analysis

View full text Add to dashboard Cite

Abstract-Parallelization and locality optimization of affine loop nests has been successfully addressed for shared-memory machines. However, many large-scale simulation applications must be executed in a distributed-memory environment, and use irregular/sparse computations where the control-flow and arrayaccess patterns are data-dependent.In this paper, we propose an approach for effective parallel execution of a class of irregular loop computations in a distributedmemory environment, using a combination of static and runtime analysis. We discuss algorithms that analyze sequential code to generate an inspector and an executor. The inspector captures the data-dependent behavior of the computation in parallel and without requiring complete replication of any of the data structures used in the original computation. The executor performs the computation in parallel. The effectiveness of the framework is demonstrated on several benchmarks and a climate modeling application.

show abstract

“…The key idea is to perform run-time analysis for predetermining the communication between processors and aggregating the messages. The notion of inspector/executor was implemented in the PARTI/CHAOS library [11,12] which was used to implement efficient and scalable message passing versions of several irregular applications like the EULER solver [13] from which IRREG is derived, and the molecular dynamics code CHARMM [14].…”

Section: Related Workmentioning

confidence: 99%

Porting and performance evaluation of irregular codes using OpenMP

Hisley

Agrawal

Satya-narayana

et al. 2000

Concurrency: Pract. Exper.

View full text Add to dashboard Cite

SUMMARYIn the last two years, OpenMP has been gaining popularity as a standard for developing portable shared memory parallel programs. With the improvements in centralized shared memory technologies and the emergence of distributed shared memory (DSM) architectures, several medium-to-large physical and logical shared memory configurations are now available. Thus, OpenMP stands to be a promising medium for developing scalable and portable parallel programs.In this paper, we focus on evaluating the suitability of OpenMP for developing scalable and portable irregular applications. We examine the programming paradigms supported by OpenMP that are suitable for this important class of applications, the performance and scalability achieved with these applications, the achieved locality and uniprocessor cache performance and the factors behind imperfect scalability. We have used two irregular applications and one NAS irregular code as the benchmarks for our study. Our experiments have been conducted on a 64-processor SGI Origin 2000.Our experiments show that reasonably good scalability is possible using OpenMP if careful attention is paid to locality and load balancing issues. Particularly, using the Single Program Multiple Data (SPMD) paradigm for programming is a significant win over just using loop parallelization directives. As expected, the cost of remote accesses is the major factor behind imperfect speedups of SPMD OpenMP programs.

show abstract

Execution time support for adaptive scientific algorithms on distributed memory machines

Cited by 55 publications

References 9 publications

Analyses and Optimizations for Shared Address Space Programs

Analyses and Optimizations for Shared Address Space Programs

Code generation for parallel execution of a class of irregular loops on distributed memory systems

Porting and performance evaluation of irregular codes using OpenMP

Contact Info

Product

Resources

About