In: Computer Science
In operating system:
How can a multicore processor take advantage of data parallelism?
Data parallelism is a programming technique in which large data set is split into smaller chunks so that we can operate on them paralelly.Once the data is processed,it is combined into a single data set.This characteristic of data parallelism is used in multicore processor to achieve better use of the the multiple cores as a process that would not be capable of utilizing multicore processing power can effiiciently use all the available processing power.
Ex-1 Consider a scenario in which a large data is being operated on by a single processor though there are 3 other CPU cores.These cores are sitting idle while the first processor has to bear the load of the whole data set
Ex-2 Now consider if we use data parallelism in order to fully use the processing power offered by a quad-core multiprocessor,we can divide the large data set into four subsets.Then each of these subsets can be individually assigned to the available four cores for processing.Once the processing is complete they can be rejoined to create a single data set.