<div dir="ltr"><div class="gmail_quote" style="font-size:12.8px">Eric:</div><div class="gmail_quote" style="font-size:12.8px">The nonscalable implementation is robust, and faster for small to medium size problems, thus we set it as the default. You can switch with option '<span style="color:rgb(80,0,80);font-size:12.8px">-matmatmult_via scalable', which requires estimate of nonzeros A*B.</span></div><div class="gmail_quote" style="font-size:12.8px"><font color="#500050">The estimate was buggy, not well-tested. If you encounter any problem, let us know.</font></div><div class="gmail_quote" style="font-size:12.8px"><font color="#500050"><br></font></div><div class="gmail_quote" style="font-size:12.8px"><font color="#500050">Hong</font></div></div><div class="gmail_extra"><br><div class="gmail_quote">On Mon, Jun 20, 2016 at 10:49 PM, Eric Chamberland <span dir="ltr"><<a href="mailto:Eric.Chamberland@giref.ulaval.ca" target="_blank">Eric.Chamberland@giref.ulaval.ca</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class=""><br>
<br>
Le 2016-06-20 23:37, Barry Smith a écrit :<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
On Jun 20, 2016, at 10:32 PM, Eric Chamberland <<a href="mailto:Eric.Chamberland@giref.ulaval.ca" target="_blank">Eric.Chamberland@giref.ulaval.ca</a>> wrote:<br>
<br>
ok, but what about -matmatmult_via scalable?<br>
</blockquote>
Both should work. It just may be one is faster or slower than the other depending on the problem size.<br>
</blockquote></span>
ok, digging further, I found this into blaming the code:<br>
<br>
0fc8cf34 (Hong Zhang 2013-06-27 14:04:58 -0500 696) /* same as MatMatMultSymbolic_MPIAIJ_MPIAIJ_nonscalable(), except using LLCondensed to avoid O(BN) memory requirement */<br>
<br>
But the commit comment says:<br>
...<br>
rename MatMatMultSymbolic_MPIAIJ_MPIAIJ -> MatMatMultSymbolic_MPIAIJ_MPIAIJ_nonscalable (non-default)<br>
<br>
But it *is* the default... since another commit:<br>
<br>
commit 0d3441ae8a080c728abf17e90308c510e39e951b<br>
Author: Hong Zhang <<a href="mailto:hzhang@mcs.anl.gov" target="_blank">hzhang@mcs.anl.gov</a>><br>
Date: Mon Aug 24 16:40:35 2015 -0500<br>
<br>
add MatPtAPxxx_MPIAIJ_MPIAIJ_new<br>
<br>
which changed the behaviour programmed in 0fc8cf34. Is it normal?<span class="HOEnZb"><font color="#888888"><br>
<br>
Eric<br>
<br>
<br>
</font></span></blockquote></div><br></div>