Automatic parallelization: Difference between revisions

Revision as of 12:41, 7 October 2006

Revision as of 19:24, 30 November 2006

Line 32:

==Workaround ==

Due to the inherent difficulties in full automatic parallelization, several easier approaches exist to get a parallel program in higher quality. They are:

* use [http://www.pervasivedatarush.com DataRush] if the application is highly data-intensive

* let programmers to add "hints" into programs to guide compiler parallelization, such as [[HPF]] for [[distributed memory]] systems and [[OpenMP]] for [[shared memory]] systems.

* build an interactive system between programmers and parallelizing tools/compilers. Notable examples are [[SUIF]] explorer, [[Polaris compiler]], and [[CAPTOOLS]](now [[Parawise]]).