Loop-level parallelism: Difference between revisions

Revision as of 00:43, 16 September 2016

Revision as of 00:47, 16 September 2016

Line 195:

=== DOACROSS parallelism ===

DOACROSS Parallelism exists where iterations of a loop are parallelized by extracting calculations that can be performed independently and running them simultaneously<ref>{{cite book|last1=Unnikrishnan|first1=Priya|title=A Practical Approach to DOACROSS Parallelization|pages=219–231|url=http://download.springer.com/static/pdf/647/chp%253A10.1007%252F978-3-642-32820-6_23.pdf?originUrl=http%3A%2F%2Flink.springer.com%2Fchapter%2F10.1007%2F978-3-642-32820-6_23&token2=exp=1473797903~acl=%2Fstatic%2Fpdf%2F647%2Fchp%25253A10.1007%25252F978-3-642-32820-6_23.pdf%3ForiginUrl%3Dhttp%253A%252F%252Flink.springer.com%252Fchapter%252F10.1007%252F978-3-642-32820-6_23*~hmac=254f4a4ae181734da221ce06bcdb03ef232f25531eaf1195c6e1988a33b9d259|accessdate=13 September 2016}}</ref>. DOALL and DOACROSS parallelisms is also type of [[Data parallelism|Data Parallelism]] since different parallel tasks perform same set of computations on different set of data.<ref name="Solihin" />

Synchronization exists to enforce loop-carried dependence.