AMBER: how to perform parallel MD on a linux cluster with Platform LSF product

From: Yu Zh.H. <computation.mail.nankai.edu.cn>
Date: Fri, 28 Apr 2006 18:31:08 +0800

Dear All,

    I'm a rookie in amber, I just compiled amber8 successfully on a linux cluster, the serial version was tesed completely correct, while I don't know how to use the parallel version (compiled with mpich) in our cluster. Our cluster use the Platform LSF product as job management, computations were performed on computating nodes (dual Xeon) after submitting a job script on management nodes. The head of script is as follows, now my question is how to perform parallel MD in my job script? and anything else need to be done? such as set environment variable...

    Thanks for your help and kindness!


#-----script head-------
#!/bin/bash
#BSUB -q normal
#BSUB -J jobname
#BSUB -o %J.output
#BSUB -n 8
#BSUB -R span[ptile=2]

(how to write here to perform parallel MD?)

#------------------------


               yours sincerely, Zhihong



-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Sun Apr 30 2006 - 06:07:09 PDT
Custom Search