Rutger ter Borg wrote:
Stricly speaking, yes, but from a user's point of view I don't think so. Please see my reply to Pavel.
Done and answered ;)
I would say LAPACK is where the fun starts. Selecting algorithms based on the lowest complexity would be good I guess. One could think of it as a compile-time analytic/semantic preprocessor of an expression.
It's what NT2 do.
I would say let's start where they (fortran-based code) are now, and improve from there? I.e., start with plugged BLAS and LAPACK, replace with faster C++ constructs Yes for real tricky stuff (starting with matrix-matrix product). NT2 is for example able to call the proper LAPACK function when doing a*b, a*trans(b), trans(a)*b etc ...
It's really easy to do with proper proto transform for checking composition and generating the proper call. We solved a large part of the interfacing problem by designing a template aware LAPACK bindings (http://lpp.sourceforge.net). -- ___________________________________________ Joel Falcou - Assistant Professor PARALL Team - LRI - Universite Paris Sud XI Tel : (+33)1 69 15 66 35