AMD Core Math Library

[Michael Hopkins]

Hi all

I am wanting to link BLAS/LAPACK routines to the optimised libraries
available from AMD (ACML 2.7) and Intel (MKL 7.2.1) and was wondering if
anyone else had done this successfully or had any advice for me.

I am currently using ATLAS and it works well but am considering distributing
software to various different x86(_64) platforms without having to rebuild
ATLAS in each case.

There are three distinct build/link scenarios:

1) Natively on FreeBSD amd64 (using Athlon64/Opteron, hence ACML)

2) Cross-compiling to 32-bit Linux (Intel & AMD with
/compat/linux/usr/bin/cc)

3) Cross-compiling to Win32 (Intel & AMD with /usr/local/mingw32/bin/cc)


As both the libraries are intended for Linux, I am not sure of the
implications for (1) & (3) - though the earlier replies to this thread seem
to suggest no problem for (1) at least.  Any tips on how to achieve this
would be welcomed.  Currently linking to ATLAS like this in all 3 scenarios:

   -L$(ATLASLIB) -llapack -lcblas -latlas


TIA

Michael


_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/

        _/    _/   _/_/_/             Hopkins Research Ltd
       _/    _/   _/    _/
      _/_/_/_/   _/_/_/          http://www.hopkins-research.com/
     _/    _/   _/   _/
    _/    _/   _/     _/               'touch the future'
                   
_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/