Automatic parallelization: Difference between revisions

Content deleted Content added
mNo edit summary
Workaround: Added maintenance template for tagging articles that overuse lists/tables
Line 104:
 
==Workaround==
{{Prose|date=February 2021}}
Due to the inherent difficulties in full automatic parallelization, several easier approaches exist to get a parallel program in higher quality. They are:
* Allow programmers to add "hints" to their programs to guide compiler parallelization, such as [[High Performance Fortran|HPF]] for [[distributed memory]] systems and [[OpenMP]] or [[OpenHMPP]] for [[shared memory (interprocess communication)|shared memory]] systems.