data-parallel computing

noun
UK /ˈdeɪtə ˈpærəlɛl kəmˈpjuːtɪŋ/
US /ˈdeɪtə ˈpærəlɛl kəmˈpjuːtɪŋ/

A way of doing the same job on lots of pieces of information at the same time.

The weather team used data-parallel computing to check millions of temperature readings quickly.


OTHER MEANINGS

2. A programming model where the same operation is applied to every element of a large dataset across many processors.

The library supports data-parallel computing so each core handles a slice of the array.

3. A hardware approach in which many processing units work on separate chunks of data under one instruction stream.

Graphics cards are built for data-parallel computing, running thousands of threads at once.

4. A technique in big-data systems for splitting a dataset into partitions that are processed simultaneously.

The cluster relies on data-parallel computing to train the model on billions of records.