Dear friends,<br><br>We have observed the following scaling problem on our cluster:<br><br>Cluster config:<br>Nodes = 3, Processor = Xeon quad core 1.86 Ghz, 2 processors per node.<br>Network= gigabit<br><br>We have used the "dppc" system from gromacs website to benchmark this cluster.
<br>The following are the results, with gromacs 3.3.2 ( configured with MPICH2 )<br><br>1. With np 8, we get 1.29 ns/day ( all running on 1 node)<br>2. With np 12 we get 1.50 ns/day <br>3. With np 16 we get 1.433 ns/day<br>
4. With np 24 we get 0.517 ns/day<br><br>We understand the scalability decreases with number of processors, but why the ns/day comes down with higher job distribution?<br><br>Kindly advice whats going wrong...<br>Thanks in advance...
<br><br>M.N. Manoj<br>CEO<br>Jigsaw Bio Solutions,<br>Bangalore - INDIA<br clear="all"><br>