Hadas Leonov wrote: > Great, thanks! > > It does solve the ia32 compilation problems, but not the lam-mpi > compilation problems - I still get undefined symbols for a few lam > variables. > > So still using open-mpi, Gromacs works a little better now, for d.villin > benchmark, the performance are: > 1 CPU: 13214 ps/day (compared to 3592 ps/day before, and 18000ps/day on > OSX 10.4) > 4 CPUs: 38000ps/day (compared to 13714 ps/day before, 41143ps/day on > OSX 10.4, and 48000ps/day in the gmx published benchmark). The performance differences might as well have to do with LAM's memory management. For some benchmark systems I have seen a 10% performance improve on a *single* CPU when Gromacs was compiled with LAM-MPI support compared to the non-MPI version. This was on Linux, however.
Carsten _______________________________________________ gmx-users mailing list gmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please search the archive at http://www.gromacs.org/search before posting! Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php
