Revision as of 07:50, 14 May 2017 edit 139.190.209.112 (talk) →Mesh interconnection network ← Previous edit		Revision as of 22:15, 28 May 2017 edit undo InternetArchiveBot (talk \| contribs) Bots, Pending changes reviewers 5,694,854 edits Rescuing 1 sources and tagging 0 as dead. #IABot (v1.3.2.4) Next edit →
Line 6: An example of MIMD system is [[Xeon Phi\|Intel Xeon Phi]], descended from [[Larrabee (microarchitecture)\|Larrabee]] microarchitecture.<ref>http://perilsofparallel.blogspot.gr/2008/09/larrabee-vs-nvidia-mimd-vs-simd.html</ref> These processors have multiple processing cores (up to 61 as of 2015) that can execute different instructions on different data. NVIDIA graphics cards fit the MIMD model, whereas the AMD/ATI cards more closely resemble the SIMD model, and have a larger number of simpler processors. Most parallel computers, as of 2013, are MIMD systems.<ref>{{cite web\|url=http://software.intel.com/en-us/articles/mimd \|title=Archived copy \|accessdate=2013-10-16 \|deadurl=yes \|archiveurl=https://web.archive.org/web/20131016215430/http://software.intel.com/en-us/articles/mimd \|archivedate=2013-10-16 \|df= }}</ref> ==Shared memory model==

Multiple instruction, multiple data: Difference between revisions