Hello,
Sometimes ago I had a problem something like this...I unphysically played with
the numbers in input file to find source of problem but nothing happened to
make me happy...I decided to observe the RAM status in graphic...the cache RAM
size was linearly(or dramatically) increasing while number of nkpt was
increasing...disk_io='low' is the default value that is to make faster
read/write speed...it means RAM is big brother...I changed input/output status
low to high (disk_io='high' in the "&CONTROL / "section of input), then
pw.x_nscf has been finished succesfully for desired numbers of nkpt. In my case
it works...I am not sure that can it solve or exactly related with your
problem?!...please just try.
Hope That Helps,
Mutlu.
> Dear
> I want to calculate optic properties of ZnO bulk using quantum
> espresso .The papers reported to need many k-points (50* 50 *31) for
> calculation optic properties of ZnO bulk .I calculate nscf using 27*
> 27*18 k_point and calculate epsilon.x. But chart is not corresponding
> of other papers ( using other cods) .
> I use 36*36*27 kpoint , but nscf stops on Computing kpt #: 19459
> Pleas guide me
> Thank you
>
------------------------------------------Dr. Mutlu ?OLAKO?ULLARITrakya
Universitesi Fen FakultesiFizik Bolumu
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://pwscf.org/pipermail/pw_forum/attachments/20140730/b46c7f32/attachment.html
