...and it can lead to parallel processing discussions for very large arrays