I think because each file is its own "task", and each task is only using one thread, it is therefore concurrent and not parallel. If it was parallel, the compression of a single file would use multiple threads. It's concurrent because multiple tasks are in progress at the same time.