Dear Amber users
We are trying to get Amber7 running well on our cluster, but are having
serious problems getting it to scale right. We've been using the DNA
tutorial as a benchmark. Here are the walltime for different sander runs
with the same .in file with differnt numbers of CPUs:
1: 1246s
2: 755s
4: 440s
8: 355s
16: 439s
As you can see, the walltime actually rises when going from 8 to 16 CPUs..
I've tested with more systems, and it seems that the more atoms that are
in the system, the worse are the scaling problems.
With 4 CPUs, each of the processors are only around 65% utilized, and
this figure drops drastically with the number of CPUs, consistant with
the walltimes.
Our cluster is a large number of intel desktop machines joined on a
gigabit ethernet. On a single machine, amber runs at the speed expected
at that clock speed. Do anyone have suggestions for improving the
performance. I have a few ideas/questions myself:
1: Would upgrading the cluster network to something with a lot smaller
latency give a better performance?
2: Is the problem inherent to Amber7, and would upgrading to Amber8 help?
Sincerely
Jacob Poehlsgaard
Dept. of Biochemistry and Molecular Biology SDU Denmark
--
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Mon Jul 12 2004 - 09:53:00 PDT
