Article: Parallelism at HotPar 2010
By: none (none.delete@this.none.com), August 13, 2010 9:10 am
Room: Moderated Discussions
AM (myname4rwt@jee-male.com) on 8/13/10 wrote:
---------------------------
[...]
>which many of us would happily
>forgive Intel if they provided the tuned versions of the code on their website --
>both in traditions of science and to promote sales of their hardware?!
That's alas the case that most (all?) papers providing
benchmark results never provide a way to reproduce them, and
this also applies to the papers linked by nVidia. BTW all
the arguments you gave that counter Intel paper apply equally
well to the nVidia-linked articles, but on one side you have a
company doing marketing, while on the other you have
"scientific" papers.
---------------------------
[...]
>which many of us would happily
>forgive Intel if they provided the tuned versions of the code on their website --
>both in traditions of science and to promote sales of their hardware?!
That's alas the case that most (all?) papers providing
benchmark results never provide a way to reproduce them, and
this also applies to the papers linked by nVidia. BTW all
the arguments you gave that counter Intel paper apply equally
well to the nVidia-linked articles, but on one side you have a
company doing marketing, while on the other you have
"scientific" papers.
