Dear All,

       in my opinion a first step could be to parallelize the SparseMatrixVector products.

On the other hand it does not make sense in my opinion to parallelize the operators over the dense matrices and vectors since this are used (at least by me) for small matrices (say 10*10) and within parallel regions.

greetings to everone

Riccardo

On Thu, 17 Nov 2011 19:01:55 +0000, David Bellot wrote:

no and yes.
So far we have no parallelization at all, apart from the implicit vectorization made by the compiler. :-(
Yes it's in the pipe but honnestly, ublas needs to be transformed a little bit. I had a look at how Eigen is doing all the vector/parallel/multicore things and we need to make some change into the code base before doing that.

Also have a look at the numeric bindings in the boost repository, they give you access to plenty of nice fast code.

David

On Thu, Nov 17, 2011 at 12:04, Kraus Philipp <philipp.kraus@flashpixx.de> wrote:

Hello,

I'm using a lot of Boost UBlas structures. Does the UBlas structures use OpenMP? Is there a support in the future?

Thanks

Phil
_______________________________________________
ublas mailing list
ublas@lists.boost.org
http://lists.boost.org/mailman/listinfo.cgi/ublas
Sent to: david.bellot@gmail.com