NWChem computation: multi-CPU run never converges


Click here for full thread
Clicked A Few Times
Dear Bert,

Quote:Bert Aug 3rd 9:23 am
Just ran it with 16 processors. I would remove the "io replicated" from the input (that's what I did).


thanks a lot! That works!!! :-)

May I ask you for short info, what does the option "io replicated" influence? And why it brokes such type of computations? And when it is (or a different io scheme) suitable to use?

Quote:Bert Aug 3rd 9:23 am
Also, I seriously doubt you can build one binary that will work on all platforms. Also, we have never tested running MPI-SPAWN between heterogeneous nodes on different clusters.


I see. But our clusters are not of different platforms -- all our clusters are "x86_64" architecture (having as much as possible unique OS/SW equipment), and thus I think that such a binary should work for us. And if not, I'll solve it once such a problem appears...

Once again: THANK YOU VERY MUCH FOR YOUR HELP AND SPENT TIME.

--best
Tom.