A Parallel Computational Model for Heterogeneous Clusters

Bosque, José Luis; Pastor, Luis

doi:10.1109/tpds.2006.165

Cited by 17 publications

(17 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There are different scalability and efficiency indicators for heterogeneous systems referred in bibliography [1,3,5]. Many references suggest that the node with more computational power should be taken as a reference.…”

Section: Discussionmentioning

confidence: 99%

“…As a result, a hybrid parallel programming model has been applied to exploit the two levels of parallelism, inside each node (with POSIXThreads) and outside each node (with OpenMPI). Each heterogeneous system (HS-X, X ∈ {A, B, C, D}) is composed of nodes (1)(2)(3)(4)(5). Each node has a Maximum Number of Cores given by MNC.…”

Section: Discussionmentioning

confidence: 99%

“…Each node has a Maximum Number of Cores given by MNC. In Table 1 the set of four heterogeneous systems used in our experiments is described, where each row specifies the number of cores used from each node (1)(2)(3)(4)(5). It means that the data in Table 1 gives the number of used cores of each node in each heterogeneous system.…”

Section: Discussionmentioning

confidence: 99%

“…For a heterogeneous system this condition means that the workload of every node has to be proportional to its computational power or speed [1,3]. Then, to optimize the parallel implementation of an algorithm on heterogeneous multiprocessors, it is necessary to know: (1) the computational power or speed of each node of multiprocessor; (2) a method for decomposing the problem into smaller subproblems (data partition/distribution); and moreover (3) data dependencies among subproblems and synchronization points.…”

Section: The Keys Of Adithementioning

confidence: 99%

“…The rest of Algorithm 2 is similar to Algorithm 1, just including the necessary communication points to develop its parallel implementation. Algorithm 2 includes the following communication points: (CP#1) The input image f is scattered across all nodes homogeneously; (CP#2) All CPU_Times measured in stage 1 are gathered to node 1, moreover, the updated image from the previous processing stages is also gathered to node 1; (CP#3) Node 1 scatters the updated image according to the values of α j and the balance condition (1); (CP#4 and CP#5) These communication points are necessary to interchange the halo data between neighboring nodes due to the data Algorithm 2 Parallel Algorithm using ADITHE; (j refers to every node j = 1 to P ) 1 f ← (f K 3 )(x, y) ∀(x, y) ∈ Sub-image j {Halo data interchange (CP#4)} 10: end for 11: for i = 1 to 7 do {Dilation Process} 12:…”

Section: Case Study: Hyperspectral Imagingmentioning

confidence: 99%

See 4 more Smart Citations

Automatic tuning of iterative computation on heterogeneous multiprocessors with ADITHE

et al. 2009

View full text Add to dashboard Cite

This work studies the problem of balancing the workload of iterative algorithms on heterogeneous multiprocessors. An approach, called ADITHE, is proposed and evaluated. Its main features are: (1) using a homogeneous distribution of the workload on the heterogeneous system, the speed of every node is estimated during the first iterations of the algorithm; (2) according to the speed of every node, a new workload distribution is carried out; (3) the remaining iterations of the algorithm are executed. The result of this workload redistribution is that the execution times for every iteration at every node are similar and, consequently, the penalties due to synchronization between nodes at every iteration are mostly eliminated. This approach is appropriate for iterative algorithms with similar workload at every iteration, and with a relevant number of iterations. The high portability of ADITHE is guaranteed because the estimation of speed of nodes is included in the execution of the parallel algorithm. There is a wide variety of iterative algorithms related to science and engineering which can take advantage of ADITHE. An example of this kind of algorithms (morphological processing of hyperspectral images) is considered in this work to evaluate its performance when ADITHE is applied. The analysis of the re-152 J. A. Martínez et al. sults shows that ADITHE significantly improves the performance of parallel iterative algorithms on heterogeneous platforms.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: The Keys Of Adithementioning

confidence: 99%

Section: Case Study: Hyperspectral Imagingmentioning

confidence: 99%

See 3 more Smart Citations

Automatic tuning of iterative computation on heterogeneous multiprocessors with ADITHE

et al. 2009

View full text Add to dashboard Cite

show abstract

Performance evaluation of model‐driven partitioning algorithms for data‐parallel kernels on heterogeneous platforms

Rico‐Gallego

Díaz‐Martín

Moreno‐Álvarez

et al. 2019

Comp and Math Methods

View full text Add to dashboard Cite

Resource analyses for parallel and distributed coordination

Trinder

Cole

Hammond

et al. 2011

Concurrency and Computation

View full text Add to dashboard Cite

SUMMARYPredicting the resources that are consumed by a program component is crucial for many parallel or distributed systems. In this context, the main resources of interest are execution time, space and communication/synchronisation costs. There has recently been significant progress in resource analysis technology, notably in type-based analyses and abstract interpretation. At the same time, parallel and distributed computing are becoming increasingly important.This paper synthesises progress in both areas to survey the state-of-the-art in resource analysis for parallel and distributed computing. We articulate a general model of resource analysis and describe parallel/distributed resource analysis together with the relationship to sequential analysis. We use three parallel or distributed resource analyses as examples and provide a critical evaluation of the analyses. We investigate why the chosen analysis is effective for each application and identify general principles governing why the resource analysis is effective.

show abstract

A Parallel Computational Model for Heterogeneous Clusters

Cited by 17 publications

References 28 publications

Automatic tuning of iterative computation on heterogeneous multiprocessors with ADITHE

Automatic tuning of iterative computation on heterogeneous multiprocessors with ADITHE

Performance evaluation of model‐driven partitioning algorithms for data‐parallel kernels on heterogeneous platforms

Resource analyses for parallel and distributed coordination

Contact Info

Product

Resources

About