Study on Template for Parallel Computing in Visual Parallel Programming Platform

With the increasing requirements of computing in a variety of fields, high performance computing has been becoming more and more popular. However, user environment of parallel computing is a bottleneck with the fast advancement of computer technology. To design and implement a convenient parallel pr...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:2006 IEEE 1st International Symposium on Pervasive Computing and Applications : August 3-5, 2006 Urumchi, Xinjiang, P.R. China s. 476 - 481
Hlavní autoři: Xiao-ping Liu, En-zhu Wang, Li-ping Zheng, Xing-wu Wei
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 01.08.2006
Témata:
Application software
Concurrent computing
High performance computing
Libraries
Message passing
parallel computing
Parallel processing
Parallel programming
Pervasive computing
Programming profession
template
visual parallel programming environment
Workstations
ISBN:1424403251, 9781424403257
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:With the increasing requirements of computing in a variety of fields, high performance computing has been becoming more and more popular. However, user environment of parallel computing is a bottleneck with the fast advancement of computer technology. To design and implement a convenient parallel programming platform for Windows environment to satisfy users, especially non-professional users, a parallel programming platform, called VPPE (Visual Parallel Programming Environment) that employs templates is described. The paper summarizes research on template for parallel computing in visual parallel programming platform in two aspects: parallel pattern templates and parallel code template. By two level templates, the platform can generate a whole parallel program in MPI+C or MPI+FORTRAN by virtue of graphs and icons provided in VPPE. Parallel mechanism is hidden by graphs and icons, which lightens user's workload and makes parallel programming easy to a certain extent
ISBN:1424403251
9781424403257
DOI:10.1109/SPCA.2006.297439