Category:Massively parallel computers

From Infogalactic: the planetary knowledge core
Jump to: navigation, search

For some time in the 1970 through 1990s, the term massively parallel was used to refer to a certain class of supercomputers. To be included, the machines had to include dozens to hundreds of individual processors, typically with their own local memory. The canonical example of a massively parallel design is the Connection Machine series.

Today, such a machine can be built using commodity hardware, an example being the System X. Many commercial systems, like Google, are based on similar designs. Modern GPUs would also be considered massively parallel by the definitions of the 1980s. For all of these reasons, the term is no longer widely used. This list is primarily concerned with earlier examples of machines that were built when parallel support was uncommon.