<div dir="ltr"><div class="gmail_extra"><div class="gmail_quote">On Tue, Apr 4, 2017 at 9:10 PM, Jed Brown <span dir="ltr"><<a href="mailto:jed@jedbrown.org" target="_blank">jed@jedbrown.org</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="">Barry Smith <<a href="mailto:bsmith@mcs.anl.gov">bsmith@mcs.anl.gov</a>> writes:<br>
<br>
> These results seem reasonable to me.<br>
><br>
> What makes you think that KNL should be doing better than it does in comparison to Haswell?<br>
><br>
> The entire reason for the existence of KNL is that it is a way for<br>
> Intel to be able to "compete" with Nvidia GPUs for numerics and<br>
> data processing, for example in the financial industry. By<br>
> "compete" I mean convince gullible purchasing agents for large<br>
> companies to purchase Intel KNL systems instead of Nvidia GPU<br>
> systems. There is nothing in the hardware specifications of KNL<br>
> that would indicate that it should work better on this type of<br>
> problem than Haswell, in fact the specifications indicate that the<br>
> Haskell should perform better<br>
<br>
</span>Boom! Time to rewrite PETSc in Haskell!<br></blockquote><div><br></div><div>Yeah, forget this debate about using C++! <br></div></div><br></div></div>