Scatter/Gather

Introduction

The MPI_Scatter primitive is sends chunks of data stored in an array to several process ranks, where each rank receives different data elements. It is similar to MPI_Broadcast, except that MPI_Broadcast sends the same data to process ranks.

The syntax is as follows:

int MPI_Scatter(const void *sendbuf, 
		int sendcount,
		MPI_Datatype sendtype,
		void *recvbuf,
		int recvcount,
		MPI_Datatype recvtype,
		int root,
		MPI_Comm comm)

The MPI_Gather primitive receives data at the root from other processes. The syntax is as follows:

int MPI_Gather(const void *sendbuf, 
		int sendcount, 
		MPI_Datatype sendtype,
		void *recvbuf, 
		int recvcount, 
		MPI_Datatype recvtype,
		int root, 
		MPI_Comm comm)

Connection to Modules

• This module may be useful for the distance matrix and K-means modules.

Files Required

• scatter_starter.c

Sum of an Array Part 1: Implementing Scatter and Gather using Send and Recv

In this exercise, rank 0 will generate an array of size $N$. There are $p$ process ranks. We assume $N~\mathrm{mod}~p=0$, i.e., $N/p$ divides evenly.

We will compute the total sum of the values in the array of size $N$ using the $p$ ranks. Using MPI_Send you will send chunks of size $N/p$ to each process rank from rank 0. Each rank will add the values in the array together ($N/p$ elements) and will send the data back to rank 0 using MPI_Send. At the end, rank 0 will add all of the values from each of the ranks together and output the total sum.

In this exercise, we will only use MPI_Send and MPI_Recv.

To get started, use the file scatter_starter.c.

Figure 1 shows an example of the sending the data, where $N=12$ and $p=4$. Each rank is sent $N/p=3$ elements to compute the local sums.

Figure 2 shows an example of sending the data from each rank to rank 0.

Sum of an Array Part 2: Using Scatter and Gather

Copy your code from Part 1. This time, implement the program using MPI_Scatter and MPI_Gather without using MPI_Send or MPI_Recv.