Abstract
We investigate opportunities for parallel execution on PowerXCell 8i processors of numerical integration algorithm as it is used in finite element codes. Different kinds of problems and different types of finite element approximations are taken into account. We design and implement several parallelization strategies and test their performance for different situations. The results of analyses and tests can be used as a guideline for choosing the proper parallelization strategy for different problems and finite element discretizations. © 2010 Springer-Verlag Berlin Heidelberg.
Cite
CITATION STYLE
Kruzel, F., & Banaś, K. (2010). Finite element numerical integration on PowerXCell processors. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 6067 LNCS, pp. 517–524). https://doi.org/10.1007/978-3-642-14390-8_54
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
