Map (parallel pattern)

Map is an <a href="/facts/Programming_idiom/jgggjC2n">idiom</a> in <a href="/facts/Parallel_computing/nQzcpzQt">parallel computing</a> where a simple operation is applied to all elements of a sequence, potentially in parallel. It is used to solve <a href="/facts/Embarrassingly_parallel/phT66Up4">embarrassingly parallel</a> problems: those problems that can be decomposed into independent subtasks, requiring no communication/synchronization between the subtasks except a <a href="/facts/Fork%E2%80%93join_model/ezh8xNer">join</a> or <a href="/facts/Barrier_(computer_science)/DUdsRXEP">barrier</a> at the end.
When applying the map pattern, one formulates an elemental function that captures the operation to be performed on a data item that represents a part of the problem, then applies this elemental function in one or more <a href="/facts/Thread_(computing)/hh8WHyPd">threads of execution</a>, <a href="/facts/Hyperthread/hjqGgK8V">hyperthreads</a>, <a href="/facts/SIMD_lanes/gQHWQSpo">SIMD lanes</a> or on <a href="/facts/Distributed_computing/o5nyhqTF">multiple computers</a>.
Some parallel programming systems, such as <a href="/facts/OpenMP/69IpI80o">OpenMP</a> and <a href="/facts/Cilk/nIftzSWW">Cilk</a>, have language support for the map pattern in the form of a parallel for loop; languages such as <a href="/facts/OpenCL/ns8y6np7">OpenCL</a> and <a href="/facts/CUDA/fvPpolot">CUDA</a> support elemental functions (as "<a href="/facts/Compute_kernel/BxyI4vSf">kernels</a>") at the language level. The map pattern is typically combined with other parallel design patterns. For example, map combined with category reduction gives the <a href="/facts/MapReduce/KbfxJ5vV">MapReduce</a> pattern.: 106–107

Map (parallel pattern) open-in-new

Map (parallel pattern)