[Wien] calculation performance

[Peter Blaha]

For a system with small H-spheres, a reasonable start for RKmax is 3-3.5.
A "very good" calculation may use 4-4.5.   (H-wavefunctions converge
very quickly with RKMAX, and for C the "effective" RKmax is twice as large
(since its RMT is twice as large), i.e. 6-7  or 8-9, which is more than enough).

In addition: Use iterative diagonalization ! A C-H system with its "few" electrons/atom
should be an ideal case for that. I expect a speedup of ~5 for such a system.

HyunJungKim schrieb:
> 
> 
> Dear users,
> 
> lapw1 runs too slow for carbon surface.
> 
> Unitcell contains 18 carbon atoms and 2 hydrogen atoms. Volume is about 
> 1200 Au^3.
> 
> 
> WFFIL        (WFPRI, SUPWF)
>   5.50       10    4 (R-MT*K-MAX; MAX L IN WF, V-NMT
>   0.30    3  0      (GLOBAL E-PARAMETER WITH n OTHER CHOICES, global 
> APW/LAPW)
>  0   -0.71      0.002 CONT 1
>  0    0.30      0.000 CONT 1
>  1    0.30      0.000 CONT 1
> ..
> 
> ..
> 
> ..
> 
> ..
> 
> This is case.in1c file. As it shows, RKmax is 5.5
> 
> 
>     At first time, I had set RKmax as 4.0 and the calculation takes 
> 13~25 hour for 1 cycle (lapw1). <-(In this case Rmt = 1.27 for C, 0.69 
> for H)
> 
>      After I change RKmax to 5.5, it takes about 2 days (50 hour). <-(in 
> this case Rmt = 1.35 for C, 0.72 for H)
> 
>     What should I do before run_lapw or runsp_lapw to minimize 
> calculation time??
> 
>     Au 111 system (similar volume about 1100 Au^3, larger RKmax 7,Rmt 
> 2.5 but 5 atoms in unitcell ) takes only few minutes for lapw1.
> 
> 
> note.) machine : Intel quad core 2.66. Single thread job ( no parallel 
> calc. )
> 
> 
> please help me.
> 
> 
> 
> 
> 
> 
> Hyun Jung, Kim
> Solid State Physics Theory Lab.
> Department of Physics, Hanyang University
> Haengdang-dong, Seongdong-gu, Seoul, South Korea
> Tel: 010-7335-7889
> email: angpangmokjang at hanmail.net
> 
> <mailto:angpangmokjang at hanmail.net>
> 
> [한메일 Express]메일목록과 미리보기를 동시에! 
> <http://allim.daum.net/servlet/Redirect?sid=footer_090223_ex>
> 	<http://daumdirect.daum.net/websales/main/event/20080516_woori_event.jsp?_partner_code.key=9100009>
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien

-- 

                                      P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-15671             FAX: +43-1-58801-15698
Email: blaha at theochem.tuwien.ac.at    WWW: http://info.tuwien.ac.at/theochem/
--------------------------------------------------------------------------