|
|
Ublas : |
Subject: Re: [ublas] Performance woes affecting ublas
From: Vardan Akopian (vakopian_at_[hidden])
Date: 2010-05-13 03:17:27
On Tue, May 11, 2010 at 3:51 PM, Rui Maciel <rui.maciel_at_[hidden]> wrote:
>
> Riccardo Rossi wrote:
> > Hi Rui,
> > this is just to say that we have a pretty large finite element
> > code heavily based on ublas ... and that in our benchmarks it is
> > competitive to fortran based implementations.
> >
> > although i am pretty sure that for small dense matrices other libraries
> > are far more efficient, i am pretty sure that this performance penalty
> > does not sussist for large sparse matrices.
>
> Is it possible to provide an example where ublas offers a reasonable
> efficiency
> at solving systems of linear equations based on sparse matrices? It would
> be
> great if it was possible to access this sort of info as the documentation
> doesn't cover this subject and is a bit lacking in these sort of details.
>
>
Hi Rui,
How about posting the code that was taking 6 minutes with uBlas and now
takes 5 seconds with eigen.
Without an example of code where library X significantly outperforms uBlas,
this thread has no value and is borderline trolling.
Otherwise, if you are actually asking for help in optimizing your uBlas
based code, you should choose a more appropriate subject line.
As to examples for using uBlas to solve sparse systems, you can do a search
on umfpack bindings for uBlas. Documentation for v1 bindings:
http://boost-sandbox.cvs.sourceforge.net/*checkout*/boost-sandbox/boost-sandbox/libs/numeric/bindings/umfpack/doc/index.html
-Vardan