INTEGRATING USER KNOWLEDGE WITH INFORMATION FROM PARALLELIZATION TOOLS TO FACILITATE THE AUTOMATIC-GENERATION OF EFFICIENT PARALLEL FORTRANCODE

Citation
Pf. Leggett et al., INTEGRATING USER KNOWLEDGE WITH INFORMATION FROM PARALLELIZATION TOOLS TO FACILITATE THE AUTOMATIC-GENERATION OF EFFICIENT PARALLEL FORTRANCODE, Parallel computing, 22(2), 1996, pp. 259-288
Citations number
18
Categorie Soggetti
Computer Sciences","Computer Science Theory & Methods
Journal title
ISSN journal
01678191
Volume
22
Issue
2
Year of publication
1996
Pages
259 - 288
Database
ISI
SICI code
0167-8191(1996)22:2<259:IUKWIF>2.0.ZU;2-B
Abstract
User supplied knowledge and interaction is a vital component of a tool kit for producing high quality parallel implementations of scalar FORT RAN numerical code. In this paper we consider the necessary components that such a parallelisation toolkit should possess to provide an effe ctive environment to identify, extract and embed user relevant user kn owledge. We also examine to what extent these facilities are available in leading parallelisation tools; in particular we discuss how these issues have been addressed in the development of the user interface of the Computer Aided Parallelisation Tools (CAPTools). The CAPTools env ironment has been designed to enable user exploration, interaction and insertion of user knowledge to facilitate the automatic generation of very efficient parallel code. A key issue in the user's interaction i s control of the volume of information so that the user is focused on only that which is needed. User control over the level and extent of i nformation revealed at any phase is supplied using a wide variety of f ilters. Another issue is the way in which information is communicated. Dependence analysis and its resulting graphs involve a lot of sophist icated rather abstract concepts unlikely to be familiar to most users of parallelising tools. As such, considerable effort has been made to communicate with the user in terms that they will understand. These fe atures, amongst others, and their use in the parallelisation process a re described and their effectiveness discussed.