Auto-generation of Parallel Finite-Differencing Code for MPI, TBB and CUDA

Playne, Daniel P.; Hawick, K. A.

doi:10.1109/ipdps.2011.265

Cited by 4 publications

(5 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…We are only just beginning to make use of domainspecific programming language techniques layered on top of our library [30]. Very modern languages with new features such as closures [11] and other boiler-plate code generation techniques [19,20] have great promise in making the underpinning libraries even more useable.…”

Section: Discussionmentioning

confidence: 99%

“…The approach we describe is also of use for managing any hyper-cubical data sets [6] that involve localised neighbour operations such as stencils [24], finite differential calculus [15,17,30] or regular pattern motif analyses. For the purposes of this present paper we focus solely on lattice simulation models [8] where each cell in the hypercubic system holds a simple data type however.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Engineering Software for d-Dimensional Simulations with Hyper-Cubic Neighbours

Hawick

2014

Software Engineering / 811: Parallel and Distributed Computing and Networks / 816: Artificial Intelligence and Applications

View full text Add to dashboard Cite

Many important simulation models can be formulated on a d-dimensional hyper-cubic lattice where the number of dimensions d is itself a parameter of the model. This leads to considerable complications even for spatially localised models as the cellular neighbourhood is a function of dimension. Neighbourhood itself can also be a necessary parameter of a model where for example different interaction distances between model degrees of freedom variables critically affects the model's emergent behaviour. We explore programming language mechanisms for coding d-dimensional data structures and simulation models and present a number of syntactic and framework approaches for engineering software with internal or external domainspecific language features.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Engineering Software for d-Dimensional Simulations with Hyper-Cubic Neighbours

Hawick

2014

Software Engineering / 811: Parallel and Distributed Computing and Networks / 816: Artificial Intelligence and Applications

View full text Add to dashboard Cite

show abstract

“…For example, in order to develop code that could run on parallel clusters automatically, Playne et al successfully developed a parallel code automatic generation system based on ANTLR compilation technology, which can automatically generate parallel programs such as MPI, CUDA, and other parallel programs written in the C/C++ languages. 26 For parallel computing, data partitioning and data flow analyses, an automatic parallel code generation module based on MPI was designed by References 27,28. Thouti et al provided a standard computing interface for heterogeneous computing platforms in order to implement an automatic parallel code on multi-core platforms.…”

Section: Related Workmentioning

confidence: 99%

Preliminary study on the automatic parallelism optimization model for image enhancement algorithms based on Intel's® Xeon Phi

Huang

Yang

Jian

et al. 2021

Concurrency and Computation

View full text Add to dashboard Cite

In unmanned aerial vehicle (UAV) image-processing applications, one needs to implement different parallel image-enhancement algorithms on several high-performance computing platforms utilizing various programming models. To speed up the parallelization procedure and improve its efficiency, the automatic parallel software package, Par4All, is applied in this work. We find that the performance of the original automatic parallelization algorithm produced with Par4All is inefficient. To resolve this problem, we propose different optimization approaches for Par4All based on Intel®'s Xeon Phi high-performance computing platform that are based on the structural features of the image-enhancement algorithms, which can further optimize the original parallel algorithm. These approaches mainly include: (1) Par4All automatic parallel search module optimization, (2) dynamic thread setting optimization, and (3) the collaborative parallelization of both CPU and many integrated core (MIC) processors.According to the results of the comparison experiments involving different algorithms, it is shown that the proposed optimization approaches for these kinds of algorithms can significantly improve the performance of automatic parallel algorithms. The acceleration ratio increases approximately by 30%, 70%, and 80% for the multiscale Retinex, Gaussian-filtering and median-filtering algorithms, respectively. As continuation and deepening of our previous research work, this research has the potential to be beneficial for other researchers in image-processing applications with image-enhancement algorithms. K E Y W O R D Sautomatic parallelism, image-enhancement algorithms, Par4All, Intel® Xeon Phi, unmanned aerial vehicles INTRODUCTIONAs mentioned by References 1, it is known that some image-enhancement algorithms, for example, median filtering, Gaussian filtering, wavelet transform, and the multiscale Retinex (MSR) algorithm, are used when preprocessing large amounts of unmanned aerial vehicle (UAV) remote-sensing (RS) images to solve the problems of poor clarity, insufficient contrast and low adaptability. In addition, many-cores computing platforms, e.g., graphics processing units (GPUs) and many integrate cores (MICs), have been involved in the development of some parallel algorithms to speed up the processing times 2 ; such approaches are expected to become mainstream in the near future. Experimental results have shown that parallelized algorithms can significantly improve computation speed and provide better acceleration efficiency. [3][4][5] Usually, there are different kinds of serial

show abstract

“…Diferentemente, a estratégia da SPar mantém a estrutura do código original e não requer conhecimento prévio do padrão de paralelismo mais adequado. Já os trabalhos de [Playne and Hawick 2011] geraram automaticamente o padrão parallel_for do TBB. Contudo, esse trabalho foi desenvolvido para aplicações matemáticas e permanece não testado em casos de stream.…”

Section: Trabalhos Relacionadosunclassified

Geração Automática de Código TBB na SPar

Filho

Griebler

Fernandes

2020

Anais Da Escola Regional De Alto Desempenho Da Região Sul (ERAD RS 2020)

View full text Add to dashboard Cite

Técnicas de programação paralela são necessárias para extrair todo o potencial dos processadores de múltiplos núcleos. Para isso, foi criada a SPar, uma linguagem para abstração do paralelismo de stream. Esse trabalho descreve a implementação da geração de código automática para a biblioteca TBB na SPar, uma vez que gerava-se código para FastFlow. Os testes com aplicações resultaram em tempos de execução até 12,76 vezes mais rápidos.

show abstract

Auto-generation of Parallel Finite-Differencing Code for MPI, TBB and CUDA

Cited by 4 publications

References 22 publications

Engineering Software for d-Dimensional Simulations with Hyper-Cubic Neighbours

Engineering Software for d-Dimensional Simulations with Hyper-Cubic Neighbours

Preliminary study on the automatic parallelism optimization model for image enhancement algorithms based on Intel's® Xeon Phi

Geração Automática de Código TBB na SPar

Contact Info

Product

Resources

About