As far as my experience goes, I've been running SIESTA on a 64-processor SMP by Hewlett-Packard. Everything was okay, although using MPI on an SMP is certainly a waste of memory. There's indeed no difference for SIESTA between an SMP and a cluster.
2007/7/19, Marcos Verissimo Alves <[EMAIL PROTECTED]>:
Hi Thomas, More details, such as compiler, mpi flavor, libraries, etc., would definitely be useful. However, I have been using siesta on my dual-core laptop without problems. Not as impressive as a 16-processed machine, but it's a SMP machine anyway :) I remember I had trouble using lam-mpi, but I don't remember if it was at compile time or run time. When I switched to mpich, everything worked fine. Regarding compiling instructions, there is no difference, as far as I know, from compilation in a cluster or smp machines. Any of the how-to's available will do the job. Marcos > Hello all,Our research group has recently invested in an SMP machine with > 16 processors. I am attempting to run compile and run the parallel version > of siesta. I am able to compile the program correctly using all of the > required libraries but problems arise when attempting to delegate a job > over multiple processors using the typical mpirun command in the siesta > manual. I know that this posting is a little vague but has anyone been > able to compile and run SIESTA multiple processors housed in a single box? > If so could you please email me so we can get into specifics.Thank you > very much for your time and I look forward to your responsesThomas > SadowskiUniversity of Connecticut > _________________________________________________________________ > Missed the show? Watch videos of the Live Earth Concert on MSN. > http://liveearth.msn.com -- Dr. Marcos Verissimo Alves Post-Doctoral Fellow Condensed Matter and Statistical Physics Sector International Centre for Theoretical Physics Trieste, Italy -------- I have become so addicted to vi that I try to exit OpenOffice by typing :wq!
