In the data parallel model, many of the parallel work focus on performing operations on a data set. The data set is usually organized into a common structure, such as an array or a cube. A set of tasks work collectively on the similar data structure with each task working on a dissimilar portion of the similar data structure. Tasks perform the similar operation on their partition of work, for example, "add 3 to each array element" can be single task. In shared memory architectures, every tasks may have access to the data structure by the global memory. In the distributed memory architectures, the data structure is dividing up and data resides as "chunks" in the local memory of each task.