Parallel-Prefix-Scan

The parallel algorithm is divided into 4 steps.1.First the input array is divided into blocks equally between the total number of threads and for each block prefix scan is done serially. But each block is worked upon in parallel by the threads. 2. An array is created with as many elements as there are threads in the program. The elements in this array were the sum of elements in one block. 3.This array goes through the reduction step and post scan step. 4. After the reduction and post scan step the elements in the array are used to increment each block of input array to achieve the required prefix scan output.