[Wien] Regarding warning in energies for SCF calculations

Gavin Abo gsabo at crimson.ua.edu
Mon Oct 15 18:51:42 CEST 2012


You should first determine how much RAM your system has (e.g., 
http://www.unixtutorial.org/2008/06/physical-memory-size-in-linux/).

I estimate that a computer or cluster needs greater than 16.5 GB of RAM 
to perform your calculation.

*How RAM was estimated:*

1. In ./siteconfig, the maximum matrix size NMATMAX was "temporarily" 
changed to a large number 999999 (must be greater than the matrix size 
found later in step 5) and then lapw1 was recompiled.
2. Initialize the calculation (TAG.struct, TAG.in1, TAG.in2, and no 
spin-polarization)
3. x lapw0
4. x lapw1 -nmat_only
5. The matrix size was 164586 in the file TAG.nmat_only.

In the UG, NMATMAX=10000 ==> 1GB (real).

164568*(1 GB/10000) = 16.4568 GB
> On Sun, Oct 14, 2012 at 12:47 PM, Kondaiah Samudrala
> <konda.physics at gmail.com> wrote:
>> Dear Sir,
>> Thank you for reply. While checking grep -e WARNING *.scf, i got below
>> error.
>>
>> grep WARN *.scf
>> :WARN :      WARNING: RKmax reduced due to NMATMAX
>> :ENE  : *WARNING** TOTAL ENERGY IN Ry =        -3716.32283352
>> :WARN :      WARNING: RKmax reduced due to NMATMAX
>> :ENE  : *WARNING** TOTAL ENERGY IN Ry =        -3716.00906001
>> :WARN :      WARNING: RKmax reduced due to NMATMAX
>> :ENE  : *WARNING** TOTAL ENERGY IN Ry =        -3715.32997770
>> :WARN :      WARNING: RKmax reduced due to NMATMAX
>> :ENE  : *WARNING** TOTAL ENERGY IN Ry =        -3714.08880483
>> :WARN :      WARNING: RKmax reduced due to NMATMAX
>> :ENE  : *WARNING** TOTAL ENERGY IN Ry =        -3713.94613786
>> :WARN :      WARNING: RKmax reduced due to NMATMAX
>> :ENE  : *WARNING** TOTAL ENERGY IN Ry =        -3713.93614806
>> :WARN :      WARNING: RKmax reduced due to NMATMAX
>> :ENE  : *WARNING** TOTAL ENERGY IN Ry =        -3713.92433048
>>
>>
>> Pls help me
>>
>>
>> with regards
>> S.Appalakondaiah
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20121015/ec5664a0/attachment.htm>


More information about the Wien mailing list