[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: modeller



Hi, 

You don't need a script to split the jobs on one machine. I used it on a
two processor PIII Linux running Rad Hat and it was decided by the machine
how to use the CPUs: load was always balanced depending on the total
number/load of other jobs. Ursula Pieper probably knows much more about
this.

Thanks,
Bozidar



On Tue, 7 May 2002, Doug Kojetin wrote:

> Thanks for your reply-                                                
>                                                                       
> In using both processors for multiple jobs -- would executing the     
> "mod" command in two separate terminals efficiently use the           
> processors?  Or do you have to use a fancy script/command to have one 
> job use processor one, and the other use processor two (or multiples  
> thereof)?                                                             
>                                                                       
> Many thanks for your help,                                            
> Doug Kojetin                                                          
>                                                                       
>                                                                       
>                                                                       
> > Hi                                                                  
> >                                                                     
> >                                                                     
> > modeller is not running parallel. The only way to take advantage of
> > multiple processors is to run parallel modeller jobs. It is useful  
> if                                                                    
> > you need to calculate many models for a protein, let say, with      
> various                                                               
> > alignments, or with various templates. Running multiple modeller    
> jobs is                                                               
> > particularly useful if you do loop modeling, since you need to      
> generate                                                              
> > numerous models before you select the one with the lowest energy.   
> > Pay attention that in this case you need to change the RAND_SEED,   
> which                                                                 
> > is responsible for random number generation, otherwise you get      
> identical                                                             
> > models.                                                             
> >                                                                     
> > best  wishes,                                                       
> >                                                                     
> > Andras                                                              
> >                                                                     
> >                                                                     
> >                                                                     
> >                                                                     
> > ------ Forwarded Message                                            
> > From: Douglas Kojetin <>                     
> > Date: Sun, 05 May 2002 23:13:47 -0400                               
> > To:                            
> > Subject: using modeller w/ multiple CPUs                            
> >                                                                     
> > Hi-                                                                 
> >                                                                     
> > Does anyone know if MODELLER can take advantage of a multiple CPU   
> > system?  If so, how to set it up?                                   
> >                                                                     
> > Many thanks,                                                        
> > Doug Kojetin                                                        
> >                                                                     
> >                                                                     
> > --                                                                  
> >     ,                                                               
> > Andras Fiser, PhD            # phone: (212) 327 7216                
> > The Rockefeller University   # fax:   (212) 327 7540                
> > Box 270, 1230 York Avenue    # e-mail:        
> > New York, NY 10021-6399, USA # http://salilab.org/~andras
> >                                                                     
>