<html><DIV class=eow_rm id=eow_rm style="PADDING-RIGHT: 1px; PADDING-BOTTOM: 15px; MARGIN-LEFT: 19px; WIDTH: 830px; MARGIN-RIGHT: 19px">
<TABLE cellSpacing=0 cellPadding=0 border=0>
<TBODY>
<TR>
<TD vAlign=top>hello,
<DIV><BR></DIV>
<DIV>when i run my mpi code with "mpiexec -n 1 ./myprogram", everything work fine.</DIV>
<DIV><BR></DIV>
<DIV>however, when i rn with "mpiexec -n 4 ./myprogram", performance of my program dropped significantly. </DIV>
<DIV><BR></DIV>
<DIV>I code my program to work only for "process 0" like this:-</DIV>
<DIV> if(id=0){ do_computation_task(); } else { //do_nothing };</DIV>
<DIV><BR></DIV>
<DIV>Is this mean physical resources are shared when I spawn more than one process in one physical machine(even all of work were done by "process 0" and other processes do nothing)?</DIV>
<DIV><BR></DIV>
<DIV>If the answer is "yes", can you pls tell me if it's stated somewhere in official document that I can make reference to. </DIV>
<DIV><BR></DIV>
<DIV>Thx,</DIV>
<DIV>Gra</DIV>
<DIV><BR>=================================</DIV>
<DIV> </DIV>
<DIV>It is my opinion that the described code just adds operating system overhead so that one would expect throughput to drop. It is not an MPI issue since the OS appears to be the source of the drop in performance. The OS has to manage four processes instead of one.</DIV></TD></TR></TBODY></TABLE></DIV></html>