Re: About the sense of removing -march=native (Was: Is theano worth saving?)
On Thu, Feb 09, 2017 at 02:13:24PM +0100, Jerome Kieffer wrote:
> On Thu, 9 Feb 2017 13:19:09 +0100
> Julian Taylor <jtaylor.debian@googlemail.com> wrote:
>
> > These numbers are very surprising to me.
>
> I do agree with Julian ...
> Once upon a time, I have been a computational chemist and believed I
> could get this 2-6x speed up by recompiling the whole scientific stack
> with the best compiler (intel icc & ifc) on the very specific computer.
>
> 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
Michael
Reply to: