Write the parallel matrix multiplication using $1-$D partition. For each case, complete the follows. $$ Consider a homogeneous parallel machine with the number of processors $.$ We consider only the simple case where the matrix dimension $$ is a multiple of $($i$.$e$.,/$ is an integer$).$ Implement the program using collective communications if needed, such as MPI$\_$Scatter, MPI$\_$Bcast, and MPI$\_$Gather. $$ Validate the program with $!,$# $=!,$# $=\backslash$times $+,=16,$ and $=16.$ Initialize the element of matrix value only at root $($rank$=0)$ process. Print the results and put it to the document file. $$ Analyze the results with $=1,2,4,8,16$ with n$=16$ and $256$ after initializing $!,$# $=!,$# $=1$ at root process. Measure the time, draw the speedup versus number of processors curve, and the CPU time versus the number of processors curve. Comment on what these curves mean from the point of view of scalability. $$ Provide code in C or C$++$ and add all your comments$/$analysis$/$plots into your document. $\backslash$