A proposed asynchronous object load balancing method for parallel 3d image reconstruction applications

Authors:
Jose Antonio Alvarez-Bermejo;Javier Roca-Piera
Affiliations:
Dept of Computer Architecture and Electronics, Universidad de Almería, Spain;Dept of Computer Architecture and Electronics, Universidad de Almería, Spain
Venue:
ICA3PP'10 Proceedings of the 10th international conference on Algorithms and Architectures for Parallel Processing - Volume Part I
Year:
2010

Citing 8
Cited 1

PO: an object model to express parallelism

OOPSLA/ECOOP '88 Proceedings of the 1988 ACM SIGPLAN workshop on Object-based concurrent programming
Concurrency within objects: layered approach

Information and Software Technology
CHARM++: a portable concurrent object oriented system based on C++

OOPSLA '93 Proceedings of the eighth annual conference on Object-oriented programming systems, languages, and applications
Data-parallel load balancing stategies

Parallel Computing
An Approach to Structured Parallel Programming Based on a Composition

CONIELECOMP '06 Proceedings of the 16th International Conference on Electronics, Communications and Computers
The load rebalancing problem

Journal of Algorithms
From structured to object oriented programming in parallel algorithms for 3D image reconstruction

Proceedings of the 8th workshop on Parallel/High-Performance Object-Oriented Scientific Computing
Communication-Aware Load Balancing for Parallel Applications on Clusters

IEEE Transactions on Computers

Efficient resource management for virtual desktop cloud computing

The Journal of Supercomputing

Quantified Score

Hi-index	0.00

Visualization

Abstract

Scientific applications usually exhibit irregular patterns of execution and high resource usage Parallel architectures are a feasible solution to face these drawbacks, but porting software to parallel platforms means the addition of an extra layer of complexity to scientific software Abstractions such as Object Orientation and models like the concurrent object model may be of great help to develop scientific parallel applications The shared nature of parallel architectures and the stochastic condition of parallel schedulers underline the adaptivity as a desired feature for parallel applications Load Balancers are key for achieving adaptivity, and benefit from object oriented models in issues like load migration In this paper we present our experiences when porting scientific software using the concurrent object abstraction and a method to asynchronously invoke load balancers.