Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
9656066 | Electronic Notes in Theoretical Computer Science | 2005 | 16 Pages |
Abstract
This paper introduces an approach to dynamic software composition in the context of scientific computing where high demands performance seem to prevent such flexible solutions. In our concrete however, dynamic software composition is rather a way to high-performance than an obstacle to it. We achieve this by combining dynamic architectures and task graph scheduling.
Related Topics
Physical Sciences and Engineering
Computer Science
Computational Theory and Mathematics
Authors
Jesper Andersson, Morgan Ericsson, Welf Löwe,