We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.DC

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Distributed, Parallel, and Cluster Computing

Title: Compile-time Parallelization of Subscripted Subscript Patterns

Abstract: An increasing number of scientific applications are making use of irregular data access patterns. An important class of such patterns involve subscripted-subscripts, wherein an array value appears in the index expression of another array. Even though the information required to parallelize loops with such patterns is often available in the program, present compiler techniques fall short of analyzing that information. In this paper we present a study of subscripted-subscripts, the properties that define the subscript arrays, and an algorithm based on symbolic range aggregation, that will help prove the presence of some of the properties of the subscript array in the program. We show that, in an important class of programs, the algorithm can boost the performance from essentially sequential execution to close to fully parallel.
Comments: 15 pages , 10 figures
Subjects: Distributed, Parallel, and Cluster Computing (cs.DC); Programming Languages (cs.PL)
Cite as: arXiv:1911.05839 [cs.DC]
  (or arXiv:1911.05839v1 [cs.DC] for this version)

Submission history

From: Akshay Bhosale [view email]
[v1] Wed, 13 Nov 2019 22:19:33 GMT (182kb,D)

Link back to: arXiv, form interface, contact.