Re: [ublas] matrix various matrix multiply timing results

From: Gunter Winkler (guwi17_at_[hidden])
Date: 2005-06-11 17:30:43

Am Donnerstag, 9. Juni 2005 12:18 schrieb Toon Knapen:
> I am a bit surprised. I would have expected that [RC]RC would provide
> optimal performance because (generally) the prod is calculated by
> axpy-ing rows of A with columns of B. Axpy-prod however is slowest on
> these configs. BTW: what is 'opb'?

BTW: the goto-product uses ideas from 
http://www.cs.utexas.edu/users/flame/goto/

mfg
Gunter