Re: About the sense of removing -march=native (Was: Is theano worth saving?)
On Thu, 9 Feb 2017 16:06:56 +0100
Michael Banck <mbanck@debian.org> wrote:
> > After spending a week or two of my PhD on that, I compared the performances
> > with the one from the stock debian build and it turned out to be within
> > 1% my costly build. Very disappointing to me. Good job you guys!
>
> By the way, the actual QM implementations on top of the linear algebra
> packages seem to vary much more widely in their performance, see e.g.
>
> https://github.com/mbanck/qmspeedtest
I have benchmarked (an not published) Turbomole vs Gaussian in 2008 and
the difference in performance was huge: about 8 to 10x faster for
Turbomole. Actually a single threaded Turbomole job was effectively
out-passing the parallel version of Gaussian using the full
dual-quad-core server.
With, may I am now become banned-by-gaussian :-)
Cheers,
--
Jérôme Kieffer
tel +33 476 882 445
Reply to: