TY - JOUR

T1 - Methods and tools of parallel programming

AU - Kasyanov, Victor

AU - Kasyanova, Elena

PY - 2017/1/1

Y1 - 2017/1/1

N2 - Using traditional methods, it is very difficult to develop high quality, portable software for parallel computers. In particular, parallel software cannot be developed on low cost, sequential computers and then moved to high performance parallel computers without extensive rewriting and debugging. In this paper, the CSS system being under development at the Institute of Informatics Systems is considered. The CSS is aimed to be an interactive visual environment for supporting of cloud parallel programming. The input language of the CSS system is a functional language Cloud Sisal that exposes implicit parallelism through data dependence and guarantees determinate result. The CSS system provides means to write and debug functional programs regardless target architectures on low-cost devices as well as to translate them into optimized parallel programs, appropriate to the target execution platforms, and then execute on high performance parallel computers in clouds without extensive rewriting and debugging.

AB - Using traditional methods, it is very difficult to develop high quality, portable software for parallel computers. In particular, parallel software cannot be developed on low cost, sequential computers and then moved to high performance parallel computers without extensive rewriting and debugging. In this paper, the CSS system being under development at the Institute of Informatics Systems is considered. The CSS is aimed to be an interactive visual environment for supporting of cloud parallel programming. The input language of the CSS system is a functional language Cloud Sisal that exposes implicit parallelism through data dependence and guarantees determinate result. The CSS system provides means to write and debug functional programs regardless target architectures on low-cost devices as well as to translate them into optimized parallel programs, appropriate to the target execution platforms, and then execute on high performance parallel computers in clouds without extensive rewriting and debugging.

KW - Cloud computing

KW - Computer science education

KW - Functional programming

KW - Hierarchical graph representation

KW - Parallel programming

UR - http://www.scopus.com/inward/record.url?scp=85020547598&partnerID=8YFLogxK

M3 - Conference article

AN - SCOPUS:85020547598

VL - 1839

SP - 141

EP - 154

JO - CEUR Workshop Proceedings

JF - CEUR Workshop Proceedings

SN - 1613-0073

ER -