Massively parallel

Source: Wikipedia, the free encyclopedia.
For other uses, see Massively parallel (disambiguation).

Massively parallel is the term for using a large number of

GPUs
are massively parallel architecture with tens of thousands of threads.

One approach is

BOINC, a volunteer-based, opportunistic grid system, whereby the grid provides power only on a best effort basis.[2]

Another approach is grouping many processors in close proximity to each other, as in a

interconnect becomes very important, and modern supercomputers have used various approaches ranging from enhanced InfiniBand systems to three-dimensional torus interconnects.[3]

The term also applies to

central processing units (CPUs) and random-access memory (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.[citation needed
] MPPAs are based on a software parallel programming model for developing high-performance embedded system applications.

Goodyear MPP was an early implementation of a massively parallel computer architecture. MPP architectures are the second most common supercomputer implementations after clusters, as of November 2013.[4]

Data warehouse appliances such as Teradata, Netezza or Microsoft's PDW commonly implement an MPP architecture to handle the processing of very large amounts of data in parallel.

See also

References

  1. ISBN 3-540-69261-4
    pages 1–4
  2. ISBN 3-642-10674-9
    pages 65–68
  3. ^ Knight, Will: "IBM creates world's most powerful computer", NewScientist.com news service, June 2007
  4. ^ "Archived copy". Archived from the original on 6 December 2013. Retrieved 12 January 2014.{{cite web}}: CS1 maint: archived copy as title (link)
Retrieved from "https://en.wikipedia.org/w/index.php?title=Massively_parallel&oldid=1134530479"