I can&#39;t speak to currently fair and accurate.  I do believe the ublas BLAS bindings would allow you to use GOTO or Intel MKL as a backend for serious crunching should they turn out to be optimal.<br><br>- Rhys<br><br><div class="gmail_quote">

On Mon, Aug 24, 2009 at 8:24 PM, Curtis Gehman <span dir="ltr">&lt;<a href="mailto:csgehman@mac.com">csgehman@mac.com</a>&gt;</span> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">

I&#39;m still new to uBLAS.  Today I discovered Eigen2, another C++ linear algebra library, that claims to be significantly faster.  The benchmarking don&#39;t appear to be ancient (2008-08).  It seems like Eigen2 gets most of the advantage from using SSE2 (something about which I know very little).  Is this comparison (still) fair and accurate?<br>


<br>
<a href="http://eigen.tuxfamily.org/index.php?title=Benchmark-August2008" target="_blank">http://eigen.tuxfamily.org/index.php?title=Benchmark-August2008</a><br>
<br>
Thank you.<br>
<br>
_______________________________________________<br>
ublas mailing list<br>
<a href="mailto:ublas@lists.boost.org">ublas@lists.boost.org</a><br>
<a href="http://lists.boost.org/mailman/listinfo.cgi/ublas" target="_blank">http://lists.boost.org/mailman/listinfo.cgi/ublas</a><br>
Sent to: <a href="mailto:rhys.ulerich@gmail.com">rhys.ulerich@gmail.com</a><br>
</blockquote></div><br>