[Wien] NMATMAX

Maciej Polak maciej.polak at pwr.edu.pl
Tue Jul 12 20:11:40 CEST 2016


The NMATMAX you have defined in your param.inc is the maximum size for a 
non mpi parallel calculation. If you use MPI, then the NMATMAX will be 
automatically increased by sqrt(nmpi). So if you want to increase your 
matrix size, the best way is to use mpi parallel execution. Simply 
increasing NMATMAX and recompiling lapw1 and then running it without mpi 
may lead to unrealistically high computational time.
And yes, the NMATMAX is really limiting the effective the RKmax, as you 
need more matrix elements to represent a bigger basis set.

Best regards

Maciej Polak

On 07/12/2016 07:53 PM, Luis Ogando wrote:
> Dear Wien2k community,
>
>    I am calculating a compound with H atoms. Due to this, I used RKmax 
> = 3.0 as recommended in the Wien2k homepage.
>    The calculating run without a problem, but I got :
>
> :WARN :      WARNING: RKmax reduced due to NMATMAX
>
>    In $WIENROOT/SRC_lpaw1/param.inc I have NMATMAX=13000 and, to my 
> surprise, I also have :
>
> :RKM  : MATRIX SIZE 55150LOs: 295  RKM= 2.38 WEIGHT= 1.00  PGR:
>
> when I was expecting MATRIX SIZE <= NMATMAX
>    I would like to get an effective RKmax=3.0 for testing convergence 
> and to get it I was thinking in changing NMATMAX in  
> $WIENROOT/SRC_lpaw1/param.inc and recompiling the SRC_lapw1 directory.
>    Well, as I am not sure if NMATMAX is really limiting the effective 
> RKmax, I would like to know what I misunderstood and what is the 
> correct procedure to increase the effective RKmax.
>    Thank you for your attention !
>    All the best,
>                       Luis
>
>
>
>
>
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> SEARCH the MAILING-LIST at:  http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20160712/906ea4f0/attachment.html>


More information about the Wien mailing list