By reducing inter-processor communication costs, the blocked assignment here reduces possibilities of a stall/time wasted on waiting for synchronization across cell values.