[SciPy-User] How to efficiently do dot(dot( A.T, diag(d) ), A ) ?

Hugh Perkins hughperkins@gmail....
Tue Sep 11 12:28:57 CDT 2012


On Wed, Sep 12, 2012 at 1:15 AM, Pauli Virtanen <pav@iki.fi> wrote:
> You are here bencmarking the underlying BLAS libraries. Matlab comes
> with Intel MKL, whereas your Numpy/Scipy is likely linked with ATLAS.

Hi Pauli,

Ok, that's good information.  I think that makes a lot of sense, and
sounds plausible to me.

> MKL can be faster than ATLAS, but on the other hand you can also link
> the Numpy/Scipy combination against MKL (provided you buy a license from
> Intel).

Ok, that sounds reasonable to me.

It makes me wonder though.  There is an opensource project called
'Eigen', for C++.
It seems to provide good performance for matrix-matrix multiplication,
comparable to Intel MKL, and significantly better than ublas
http://eigen.tuxfamily.org/index.php?title=Benchmark  I'm not sure
what the relationship is between ublas and BLAS?

I wonder if it might be worth scipy providing an option to link with
eigen3?  I confess I don't personally have time to do that though.


More information about the SciPy-User mailing list