Efficient parallel graph algorithms for coarse grained multicomputers and BSP

Cáceres, Edson Norberto; Dehne, Frank; Ferreira, Afonso; Flocchini, Paola; Rieping, Ingo; Roncato, Alessandro; Santoro, Nicola; Song, Siang Wun

doi:10.1007/3-540-63165-8_195

Cited by 46 publications

(25 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Additionally, there has been a considerable amount of work on algorithms design for the BSP model. For instance, sorting and related problems are considered in [17,23], list and graph problems are considered in [12], matrix multiplication and linear algebra problems are considered in [17,46,37], algorithms for dynamic data structures are considered in [8], to cite just a few. LogP Algorithmic Results.…”

Section: Bsp Algorithmic Resultsmentioning

confidence: 99%

“…Some QSM algorithmic results for sorting and list ranking that focus on reducing the number of phases are given in [43]. Related work on minimizing the number of supersteps on a BSP using the notion of rounds is reported in [23] for sorting, and in [12] for graph problems. Several lower bounds for the number of phases needed for basic problems are presented in [34].…”

Section: (Erew and Qrqw Algorithms On Qsm)mentioning

confidence: 99%

See 1 more Smart Citation

Parallel Algorithm Design with Coarse-Grained Synchronization

Ramachandran

2001

Computational Science - ICCS 2001

View full text Add to dashboard Cite

Abstract.We describe the Queuing Shared-Memory (QSM) and BulkSynchronous Parallel (BSP) models of parallel computation. The former is shared-memory and the latter is distributed-memory. Both models use the 'bulk-synchronous' paradigm introduced by the BSP model. We describe the relationship of these two models to each other and to the 'LogP' model, and give an overview of algorithmic results on these models.

show abstract

Section: Bsp Algorithmic Resultsmentioning

confidence: 99%

Section: (Erew and Qrqw Algorithms On Qsm)mentioning

confidence: 99%

Parallel Algorithm Design with Coarse-Grained Synchronization

Ramachandran

2001

Computational Science - ICCS 2001

View full text Add to dashboard Cite

show abstract

“…The list ranking problem appears frequently, see e.g Caceres et al [1997], in parallel algorithms that compute on object like lists, trees or graphs. List Ranking has a linked list of elements as input where each element knows its successor as well as the distance which separates these two elements.…”

Section: A List Rankingmentioning

confidence: 99%

An experimental validation of the PRO model for parallel and distributed computation

Essaïdi

Gustedt

2006

14th Euromicro International Conference on Parallel, Distributed, and Network-Based Processing (PDP'06)

View full text Add to dashboard Cite

Abstract-The Parallel Resource-Optimal (PRO) computation model was introduced by Gebremedhin et al. [2002] as a framework for the design and analysis of efficient parallel algorithms. The key features of the PRO model that distinguish it from previous parallel computation models are the full integration of resource-optimality into the design process and the use of a granularity function as a parameter for measuring quality. In this paper we present experimental results on parallel algorithms, designed using the PRO model, for two representative problems: list ranking and sorting. The algorithms are implemented using SSCRAP, our environment for developing coarse-grained algorithms. The experimental performance results observed agree well with analytical predictions using the PRO model. Moreover, by using different platforms to run our experiments, we have been able to provide an integrated view of the modeling of an underlying architecture and the design and implementation of scalable parallel algorithms.

show abstract

“…Related work on minimizing the number of phases (or supersteps) using the notion of rounds is reported in [13] for sorting and in [5] for graph problems. Several lower bounds for the number of rounds needed for basic problems on the QSM and BSP are presented in [19].…”

Section: Cost Measures For a Qsm Computationmentioning

confidence: 99%

“…On the other hand, a computation that proceeds in rounds need not lead to a linear work algorithm if the number of rounds in the algorithm is non-constant. In fact, all of the algorithms presented in [5] perform superlinear work. The algorithm in [13] performs superlinear communication when the number of processors is large.…”

Section: Cost Measures For a Qsm Computationmentioning

confidence: 99%

Emulations between QSM, BSP and LogP: a framework for general-purpose parallel algorithm design

Ramachandran

Grayson

Dahlin

2003

Journal of Parallel and Distributed Computing

View full text Add to dashboard Cite

We present work-preserving emulations with small slowdown between LogP and two other parallel models: BSP and QSM. In conjunction with earlier work-preserving emulations between QSM and BSP, these results establish a close correspondence between these three general-purpose parallel models. Our results also correct and improve on results reported earlier on emulations between BSP and LogP. In particular we shed new light on the relative power of stalling and non-stalling LogP models.The QSM is a shared-memory model with only two parameters-p; the number of processors, and g; a bandwidth parameter. The simplicity of the QSM parameters makes QSM a convenient model for parallel algorithm design, and simple work-preserving emulations of QSM on BSP and QSM on LogP show that algorithms designed for the QSM will also map quite well to these other models. The simplicity and generality of QSM present a strong case for the use of QSM as the model of choice for parallel algorithm design.We present QSM algorithms for three basic problems-prefix sums, sample sort and list ranking. We show that these algorithms are optimal in terms of both the total work performed and the number of 'phases' for input sizes of practical interest. For prefix sums, we present a matching lower bound that shows our algorithm to be optimal over the complete range of these parameters. We then examine the predicted and simulated performance of these algorithms. These results suggest that QSM analysis will predict algorithm performance quite accurately for problem sizes that arise in practice. r

show abstract

Efficient parallel graph algorithms for coarse grained multicomputers and BSP

Cited by 46 publications

References 23 publications

Parallel Algorithm Design with Coarse-Grained Synchronization

Parallel Algorithm Design with Coarse-Grained Synchronization

An experimental validation of the PRO model for parallel and distributed computation

Emulations between QSM, BSP and LogP: a framework for general-purpose parallel algorithm design

Contact Info

Product

Resources

About