1052:
100:(RAM) banks. These processors pass work to one another through a reconfigurable interconnect of channels. By harnessing many processors working in parallel, an MPPA chip can accomplish more demanding tasks than conventional chips. MPPAs are based on a software parallel programming model for developing high-performance embedded system applications.
378:
262:
468:
320:
288:
449:
489:
716:
739:
20:
106:
was an early implementation of a massively parallel computer architecture. MPP architectures are the second most common
628:
89:
484:
734:
711:
236:
216:
313:
128:'s PDW commonly implement an MPP architecture to handle the processing of very large amounts of data in parallel.
706:
521:
813:
727:
676:
1037:
871:
722:
409:
152:
1077:
1056:
1002:
462:
306:
157:
981:
776:
661:
623:
473:
363:
997:
976:
921:
808:
798:
771:
633:
163:
77:
becomes very important, and modern supercomputers have used various approaches ranging from enhanced
266:
951:
577:
516:
429:
113:
93:
74:
1082:
1012:
1007:
866:
457:
142:
751:
683:
587:
479:
434:
843:
803:
756:
746:
541:
404:
343:
55:
32:
783:
671:
666:
656:
643:
439:
97:
66:, opportunistic grid system, whereby the grid provides power only on a best effort basis.
8:
946:
901:
701:
567:
63:
971:
793:
618:
582:
572:
531:
373:
348:
329:
184:
174:
169:
147:
82:
69:
Another approach is grouping many processors in close proximity to each other, as in a
36:
1017:
693:
651:
546:
282:
232:
212:
35:(or separate computers) to simultaneously perform a set of coordinated computations
1027:
826:
761:
608:
424:
419:
414:
383:
189:
70:
51:
891:
831:
766:
613:
603:
536:
526:
368:
358:
209:
Grid computing: experiment management, tool integration, and scientific workflows
137:
92:(MPPAs), a type of integrated circuit with an array of hundreds or thousands of
1022:
838:
495:
388:
47:
1071:
911:
788:
107:
511:
103:
58:
is opportunistically used whenever a computer is available. An example is
1032:
78:
43:
are massively parallel architecture with tens of thousands of threads.
906:
881:
298:
125:
956:
936:
861:
117:
961:
941:
916:
551:
121:
73:. In such a centralized system the speed and flexibility of the
931:
926:
59:
248:
Knight, Will: "IBM creates world's most powerful computer",
966:
896:
886:
179:
876:
853:
40:
16:
Use of many processors to perform simultaneous operations
110:
implementations after clusters, as of
November 2013.
229:
1069:
314:
321:
307:
54:of many computers in distributed, diverse
242:
31:is the term for using a large number of
1070:
328:
287:: CS1 maint: archived copy as title (
211:by Radu Prodan, Thomas Fahringer 2007
302:
202:
231:by Francisco Fernández de Vega 2010
222:
90:massively parallel processor arrays
21:Massively parallel (disambiguation)
13:
14:
1094:
1051:
1050:
522:Analysis of parallel algorithms
255:
1:
469:Simultaneous and heterogenous
250:NewScientist.com news service
195:
81:systems to three-dimensional
1057:Category: Parallel computing
153:Process-oriented programming
7:
158:Shared-nothing architecture
131:
10:
1099:
364:High-performance computing
18:
1046:
998:Automatic parallelization
990:
852:
692:
642:
634:Application checkpointing
596:
560:
504:
448:
397:
336:
164:Symmetric multiprocessing
114:Data warehouse appliances
88:The term also applies to
94:central processing units
1013:Embarrassingly parallel
1008:Deterministic algorithm
143:Embarrassingly parallel
728:Associative processing
684:Non-blocking algorithm
490:Clustered multi-thread
56:administrative domains
844:Hardware acceleration
757:Superscalar processor
747:Dataflow architecture
344:Distributed computing
723:Pipelined processing
672:Explicit parallelism
667:Implicit parallelism
657:Dataflow programming
98:random-access memory
19:For other uses, see
947:Parallel Extensions
752:Pipelined processor
83:torus interconnects
33:computer processors
1078:Parallel computing
821:Massively parallel
799:distributed shared
619:Cache invalidation
583:Instruction window
374:Manycore processor
354:Massively parallel
349:Parallel computing
330:Parallel computing
269:on 6 December 2013
185:Manycore processor
175:Cellular automaton
170:Connection Machine
148:Parallel computing
29:Massively parallel
1065:
1064:
1018:Parallel slowdown
652:Stream processing
542:Karp–Flatt metric
1090:
1054:
1053:
1028:Software lockout
827:Computer cluster
762:Vector processor
717:Array processing
702:Flynn's taxonomy
609:Memory coherence
384:Computer network
323:
316:
309:
300:
299:
293:
292:
286:
278:
276:
274:
265:. Archived from
259:
253:
246:
240:
226:
220:
206:
190:Vector processor
71:computer cluster
52:processing power
46:One approach is
1098:
1097:
1093:
1092:
1091:
1089:
1088:
1087:
1068:
1067:
1066:
1061:
1042:
986:
892:Coarray Fortran
848:
832:Beowulf cluster
688:
638:
629:Synchronization
614:Cache coherence
604:Multiprocessing
592:
556:
537:Cost efficiency
532:Gustafson's law
500:
444:
393:
369:Multiprocessing
359:Cloud computing
332:
327:
297:
296:
280:
279:
272:
270:
263:"Archived copy"
261:
260:
256:
247:
243:
227:
223:
207:
203:
198:
138:Multiprocessing
134:
64:volunteer-based
24:
17:
12:
11:
5:
1096:
1086:
1085:
1083:Supercomputing
1080:
1063:
1062:
1060:
1059:
1047:
1044:
1043:
1041:
1040:
1035:
1030:
1025:
1023:Race condition
1020:
1015:
1010:
1005:
1000:
994:
992:
988:
987:
985:
984:
979:
974:
969:
964:
959:
954:
949:
944:
939:
934:
929:
924:
919:
914:
909:
904:
899:
894:
889:
884:
879:
874:
869:
864:
858:
856:
850:
849:
847:
846:
841:
836:
835:
834:
824:
818:
817:
816:
811:
806:
801:
796:
791:
781:
780:
779:
774:
767:Multiprocessor
764:
759:
754:
749:
744:
743:
742:
737:
732:
731:
730:
725:
720:
709:
698:
696:
690:
689:
687:
686:
681:
680:
679:
674:
669:
659:
654:
648:
646:
640:
639:
637:
636:
631:
626:
621:
616:
611:
606:
600:
598:
594:
593:
591:
590:
585:
580:
575:
570:
564:
562:
558:
557:
555:
554:
549:
544:
539:
534:
529:
524:
519:
514:
508:
506:
502:
501:
499:
498:
496:Hardware scout
493:
487:
482:
477:
471:
466:
460:
454:
452:
450:Multithreading
446:
445:
443:
442:
437:
432:
427:
422:
417:
412:
407:
401:
399:
395:
394:
392:
391:
389:Systolic array
386:
381:
376:
371:
366:
361:
356:
351:
346:
340:
338:
334:
333:
326:
325:
318:
311:
303:
295:
294:
254:
241:
221:
200:
199:
197:
194:
193:
192:
187:
182:
180:CUDA framework
177:
172:
167:
161:
155:
150:
145:
140:
133:
130:
48:grid computing
15:
9:
6:
4:
3:
2:
1095:
1084:
1081:
1079:
1076:
1075:
1073:
1058:
1049:
1048:
1045:
1039:
1036:
1034:
1031:
1029:
1026:
1024:
1021:
1019:
1016:
1014:
1011:
1009:
1006:
1004:
1001:
999:
996:
995:
993:
989:
983:
980:
978:
975:
973:
970:
968:
965:
963:
960:
958:
955:
953:
950:
948:
945:
943:
940:
938:
935:
933:
930:
928:
925:
923:
920:
918:
915:
913:
912:Global Arrays
910:
908:
905:
903:
900:
898:
895:
893:
890:
888:
885:
883:
880:
878:
875:
873:
870:
868:
865:
863:
860:
859:
857:
855:
851:
845:
842:
840:
839:Grid computer
837:
833:
830:
829:
828:
825:
822:
819:
815:
812:
810:
807:
805:
802:
800:
797:
795:
792:
790:
787:
786:
785:
782:
778:
775:
773:
770:
769:
768:
765:
763:
760:
758:
755:
753:
750:
748:
745:
741:
738:
736:
733:
729:
726:
724:
721:
718:
715:
714:
713:
710:
708:
705:
704:
703:
700:
699:
697:
695:
691:
685:
682:
678:
675:
673:
670:
668:
665:
664:
663:
660:
658:
655:
653:
650:
649:
647:
645:
641:
635:
632:
630:
627:
625:
622:
620:
617:
615:
612:
610:
607:
605:
602:
601:
599:
595:
589:
586:
584:
581:
579:
576:
574:
571:
569:
566:
565:
563:
559:
553:
550:
548:
545:
543:
540:
538:
535:
533:
530:
528:
525:
523:
520:
518:
515:
513:
510:
509:
507:
503:
497:
494:
491:
488:
486:
483:
481:
478:
475:
472:
470:
467:
464:
461:
459:
456:
455:
453:
451:
447:
441:
438:
436:
433:
431:
428:
426:
423:
421:
418:
416:
413:
411:
408:
406:
403:
402:
400:
396:
390:
387:
385:
382:
380:
377:
375:
372:
370:
367:
365:
362:
360:
357:
355:
352:
350:
347:
345:
342:
341:
339:
335:
331:
324:
319:
317:
312:
310:
305:
304:
301:
290:
284:
268:
264:
258:
251:
245:
238:
237:3-642-10674-9
234:
230:
225:
218:
217:3-540-69261-4
214:
210:
205:
201:
191:
188:
186:
183:
181:
178:
176:
173:
171:
168:
165:
162:
159:
156:
154:
151:
149:
146:
144:
141:
139:
136:
135:
129:
127:
123:
119:
115:
111:
109:
108:supercomputer
105:
101:
99:
95:
91:
86:
84:
80:
76:
72:
67:
65:
61:
57:
53:
49:
44:
42:
38:
34:
30:
26:
22:
820:
597:Coordination
527:Amdahl's law
463:Simultaneous
353:
271:. Retrieved
267:the original
257:
249:
244:
228:
224:
208:
204:
112:
104:Goodyear MPP
102:
87:
75:interconnect
68:
50:, where the
45:
28:
27:
25:
1033:Scalability
794:distributed
677:Concurrency
644:Programming
485:Cooperative
474:Speculative
410:Instruction
252:, June 2007
239:pages 65–68
96:(CPUs) and
37:in parallel
1072:Categories
1038:Starvation
777:asymmetric
512:PRAM model
480:Preemptive
273:12 January
196:References
79:InfiniBand
772:symmetric
517:PEM model
219:pages 1–4
126:Microsoft
1003:Deadlock
991:Problems
957:pthreads
937:OpenHMPP
862:Ateji PX
823:computer
694:Hardware
561:Elements
547:Slowdown
458:Temporal
440:Pipeline
283:cite web
132:See also
118:Teradata
116:such as
962:RaftLib
942:OpenACC
917:GPUOpen
907:C++ AMP
882:Charm++
624:Barrier
568:Process
552:Speedup
337:General
122:Netezza
1055:
932:OpenCL
927:OpenMP
872:Chapel
789:shared
784:Memory
719:(SIMT)
662:Models
573:Thread
505:Theory
476:(SpMT)
430:Memory
415:Thread
398:Levels
235:
215:
902:Dryad
867:Boost
588:Array
578:Fiber
492:(CMT)
465:(SMT)
379:GPGPU
166:(SMP)
60:BOINC
967:ROCm
897:CUDA
887:Cilk
854:APIs
814:COMA
809:NUMA
740:MIMD
735:MISD
712:SIMD
707:SISD
435:Loop
425:Data
420:Task
289:link
275:2014
233:ISBN
213:ISBN
160:(SN)
62:, a
41:GPUs
982:ZPL
977:TBB
972:UPC
952:PVM
922:MPI
877:HPX
804:UMA
405:Bit
124:or
1074::
285:}}
281:{{
120:,
85:.
39:.
322:e
315:t
308:v
291:)
277:.
23:.
Text is available under the Creative Commons Attribution-ShareAlike License. Additional terms may apply.