I'm sorry for being naive. Very new to mpi and parallel processing. How do I run paraview on a machine with 8 processors so that I can get the fastest performance?<br><br>Bikash<br><br><br><div><span class="gmail_quote">
On 4/2/07, <b class="gmail_sendername">Kent Eschenberg</b> <<a href="mailto:eschenbe@psc.edu">eschenbe@psc.edu</a>> wrote:</span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
Bikash Agarwal wrote:<br>> When running paraview 2.6 in MPI mode on a single machine, why does it<br>> open one instance of paraview for each processor? When I use,<br>> mpirun -np 2 ./paraview<br>> it opens up 2 instances of paraview.
<br><br>??? That's what you asked it to do with "-np 2". Each<br>instance knows it is using MPI and so uses the MPI library<br>to connect to the other instances. "mpirun" will spread the<br>instances out over as many processors as are available to it
<br>(one, in this case).<br><br>Kent<br></blockquote></div><br>