Just a bit more; for multi-cores, we should use parallel version of BLAS, LAPACK (like ATLAS, via pthread) for clusters (many machines) we should use ScaLAPACK. So we can do more with small effort :-)
I'm not sure which LP solver is the fastest. And currently lpsolve in OOo source code tree doesn't call BLAS or LAPACK. (looks like intention is mainly on mixed integer linear programming; see http://en.wikipedia.org/wiki/Linear_programming ) Thanks, -- Nakata Maho http://accc.riken.jp/maho/ , http://ja.openoffice.org/ Nakata Maho's PGP public keys: http://accc.riken.jp/maho/maho.pgp.txt --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
