You are running a program that performs a summation operation of all elements in Array A.

Input: Array A of 2048 numbers Output: Task: Parallelize this problem efficiently using multiple processing elements (PEs)

# of available PEs: 128 PEs

Assumptions: Computation - Each PE can add two numbers stored in its local memory in one second. Communication - A PE can send data from its local memory to the local memory of any other PE in three seconds (independent of the size of the data). Input and output - At the beginning of the program, the whole input array A is stored in PE #0. In the end, the result should be gathered in PE #0. Synchronization - All PEs operate in a lock-step manner; i.e., they can either compute, communicate, or be idle.

Based on the above condition, determine the best processing elements (PEs) that support the fastest runtime, speedup, and efficiency

2047 [i] i=0 2047 [i] i=0