Q. Explain Data Parallel Programming?

In data parallel programming model major focus is on performing concurrent operations on a data set. The data set is characteristically organized in a common structure like an hypercube or array. Processors work together on the same data structure. But every task works on various partition of similar data structure. It is more restrictive since not all algorithms can be specified in data-parallel terms. For these reasons data parallelism, though significant isn't a universal parallel programming paradigm.

Programming with data parallel model is generally achieved by writing a program with the data parallel constructs. The constructs can be called to a compiler directives or else data parallel subroutine library. Data parallel languages offer facilities to indicate the data decomposition and mapping to the processors. The languages comprise data distribution statements that allow the programmer to control that data goes on what processor to minimize the amount of communication between processors. Directives signify how arrays are to be aligned and distributed over processors and therefore specify agglomeration and mapping. Communications operations aren't specified explicitly by programmer however are instead inferred by compiler from program. Data parallel languages are more appropriate for SIMD architecture although some languages for MIMD structure have also been designed. Data parallel approach is more effective for highly regular problems however isn't very effective for irregular problems.