F. Baiardi, P. Mori, and L. Ricci (Italy)
Parallelization Tools; Irregular Problems; Distributed Memory Architectures; Load Balancing; Locality
This paper presents PIT, a library for the parallelization of irregular problems on distributed memory architectures. All the strategies underlying the definition of the library can be expressed in terms of operations on a PITree, a par allel version of the tree data structure oriented to irregular problems. We consider the application of PIT to two well known irregular problems: adaptive multigrid and hierar chical radiosity methods. Performance figures that prove the effectiveness of PIT are presented.
Important Links:
Go Back
