From: Axel Kohlmeyer (akohlmey_at_cmm.chem.upenn.edu)
Date: Tue May 19 2009 - 09:55:57 CDT
On Tue, 2009-05-19 at 07:29 -0500, jose correa wrote:
> Dear all
> We are trying to compile namd 2.6 on openmosix
> (http://openmosix.sourceforge.net/) using a cluster of 20 nodes, any
> suggestion or information related to it.
you should be fine just running the "net" version of NAMD
without having to recompile it. using any other communication
transport layer would just add redundant communication and
slow it down. the performance will be somewhat limited by
the interconnect latency and bandwith, but that would be
the same in a traditional cluster.
you should be able to run in parallel with charmrun using ++local
i would recommend to make some careful scaling tests so you
can give hints to the users as to how many processors they
should use for what system size. scaling will be limited and
the convenience of openmosix is paid for with additional
overhead in communication.
> Best wishes
> José Correa-Basurto
-- ======================================================================= Axel Kohlmeyer akohlmey_at_cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot.
This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:50:59 CST