<div dir="ltr"><div>"-map-by socket" was for the non-sequential ordering I had. I got sequentially ordering with "-bind-to user:0,1,2,3,4,5,6,7,10,11,12,13,14,15,16,17"<br></div></div><div class="gmail_extra"><br><div class="gmail_quote">On Mon, May 11, 2015 at 4:07 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="">Justin Chang <<a href="mailto:jychang48@gmail.com">jychang48@gmail.com</a>> writes:<br>
> ${MPIEXEC}.hydra -n <procs> -bind-to hwthread -map-by socket ./myprogram<br>
> <args><br>
<br>
</span>Why are you using "-map-by socket" if you want a sequential ordering?<br>
<br>
Abstractly, I expect a tradeoff between load balancing in case of<br>
correlated irregularity and cache locality, but would tend to prefer<br>
sequential ordering (so that nearby row numbers are more likely to share<br>
cache).<br>
</blockquote></div><br></div>