0

Hello.

I need to draw an activity digarm for a program code called Finite Differene Time Domain (FDTD). I have attached the code in a document where there are all the for loops. The task is given below

In the piece of code mentioned, you could look at parallelizing the lines of code, as I said
think as if you are using threads. Parallelized grains of code must be independent.

If you think about using a cluster (rather than a parallel machine) then you may want higher
granularity (ie you split the task into bigger chunks (since the communication times are greater).

If anybody could help me out by drwaing an activity digaram for the code below or point me towards tutorials for understanding more.

tHANKS

3
Contributors
3
Replies
10
Views
9 Years
Discussion Span
Last Post by james.wilder.9849
0

It is like matrix multiplication can be distributed among many cpu each of them can compute a row /column independently,
Nice homework I would say.

0

Any ideas as to what is happenign in the loops below. I understand ther are three loops. diffiuculy on the array and hwo to reperesent this in a dependency graph.

for (Jx = 1; Jx <= (Nx - 1); Jx++) {
for (Jy = 1; Jy <= (Ny - 1); Jy++) {
for (Jz = 1; Jz <= (Nz - 1); Jz++) {
Eold[1] = E.x.grid[Jx][Jy][Jz];
Eold[2] = E.y.grid[Jx][Jy][Jz];
Eold[3] = E.z.grid[Jx][Jy][Jz];

0

the piece of code mentioned, you could look at parallelizing the lines of code, as I said

My suggestion for you to use a online uml software since you have not specified your platform here. Loops can be generated automatically when the diagram is desinged from the scartch uml software

Edited by james.wilder.9849: lookd spammy

This topic has been dead for over six months. Start a new discussion instead.
Have something to contribute to this discussion? Please be thoughtful, detailed and courteous, and be sure to adhere to our posting rules.