000055723 001__ 55723
000055723 005__ 20180317094355.0
000055723 037__ $$aCONF
000055723 245__ $$aContention elimination by replication of sequential sections in distributed shared memory programs
000055723 269__ $$a2001
000055723 260__ $$c2001
000055723 336__ $$aConference Papers
000055723 520__ $$aIn shared memory programs contention often occurs at the transition between a sequential and a parallel section of the code. As all threads start executing the parallel section, they often access data just modified by the thread that executed the sequential section, causing a flurry of data requests to converge on that processor.We address this problem in a software distributed shared memory system by replicating the execution of the sequential sections on all processors. Communication during this replicated sequential execution is reduced by using multicast.We have implemented replicated sequential execution with multicast support in OpenMP/NOW, a version of of OpenMP that runs on networks of workstations. We do not rely on compile-time data analysis, and therefore we can handle irregular and pointer-based applications. We show significant improvement for two pointer-based applications that suffer from severe contention without replicated sequential execution.
000055723 700__ $$aLu, H
000055723 700__ $$aCox, A.L
000055723 700__ $$0243160$$aZwaenepoel, W$$g155705
000055723 7112_ $$aProceedings of the eighth ACM SIGPLAN symposium on Principles and practices of parallel programming$$cSnowbird, Utah, United States$$dJune 2001
000055723 8564_ $$s173488$$uhttps://infoscience.epfl.ch/record/55723/files/p53-lu.pdf$$zn/a
000055723 909CO $$ooai:infoscience.tind.io:55723$$pIC$$pconf
000055723 909C0 $$0252226$$pLABOS$$xU10700
000055723 937__ $$aLABOS-CONF-2005-020
000055723 973__ $$aOTHER$$rREVIEWED$$sPUBLISHED
000055723 980__ $$aCONF