Video games for example - runtime, tools, etc. We had to optimize our tool pipeline to parallelize things like dxt compression, vertex welding, mip-map creation, filters, etc. as part of the pipeline. Since there were other restrictions, that we couldn't not safely run two or more assets to be converted at the same time, we chose data parallelize (where possible) the actual asset convertion. OpenMP was used for that and helped a lot (VS2008, then VS2010).
Since the process was reading, process (data parallelization), writing - and reading was much bigger than writing (you read big model, animation, or texture files, and write ~8x smaller), it was "parallelized" by prefetching files in the file cache reading them in a second thread in advance.
Allowing only one writer to the HDD turned out to be beneficial (for cases like RAID SSD you can do multiple writers efficiently but this is rather pricey).
Since the process was reading, process (data parallelization), writing - and reading was much bigger than writing (you read big model, animation, or texture files, and write ~8x smaller), it was "parallelized" by prefetching files in the file cache reading them in a second thread in advance.
Allowing only one writer to the HDD turned out to be beneficial (for cases like RAID SSD you can do multiple writers efficiently but this is rather pricey).