Why is it faster to process a sorted array than an unsorted array in C++ program?

In C++, it is faster to process a sorted array than an unsorted array because of branch prediction. In computer architecture, a branch prediction determines whether a conditional branch (jump) in the instruction flow of a program is likely to be taken or not.

Let’s take an example:

if(arr[i] > 50) {
   Do some operation B
} else {
   Do some operation A
}

If we run this code for 100 elements in unsorted and sorted order below things will be happened:

For sorted array:

1,2,3,4,5,……50,51………100
A,A,A,A,A A,B B
It will load the correct branch in pipeline and correct sequence A,A,A, A,A,A,A,A A,B B

For unsorted array:

5,51,6,90,4,49,60…
A,B,A,B,A,A,A,B

Branch prediction doesn’t play its significant role here. It is very difficult to predict the right operation between A and B.