[Pw_forum] references for philosophy of parallel execution

Nicola Marzari marzari at MIT.EDU
Fri Sep 4 03:52:10 CEST 2009


sreekar guddeti wrote:
> hello QE users,
> i would be grateful if u could provide me with any references for the 
> principles involved in parallel execution like how the grids are 
> divided among processors which, should supposedly enhance perfomance, 
> etc. because i performed a parallel execution and the CPU time taken 
> is more than that in the serial case
>
> thnks in advance
>
>
Dear Sreekar,

you could start with the QE reference paper (Stefano Baroni just posted its
reference, freely accessible in the next 30 days on J Phys Cond Matt) and
(I would guess) ref 167 and 168 there.

If I remember correctly, a detailed explanation of parallelization in G 
space
is in

L. J. Clarke, I. Stich and M. C. Payne Large-scale ab initio total energy
calculations on parallel computers Comput. Phys. Commun. 72 14 (1992)

M. C. Payne, L. J. Clarke and I. Stich Role of parallel architectures
in periodic boudnary calculations Phil Trans. Roy. Soc. A341 211
(1992)

(the first ever electronic-structure calculations on parallel machines 
were done
on the 7x7 Takanayagi Si(111) surface, by the group of Mike Payne in
Cambridge, and separately John Joannopoulos at MIT)

I think the G space parallelization of Espresso follows the same philosophy.

Note that other parallelization schemes (read the QE article) such as 
those on
the k-points, involve very little communications, and so they should be very
efficient. More concise info on your system size you studied, and your
parallelization strategy, could allow others to give you specific advice.


             nicola

-- 
---------------------------------------------------------------------
Prof Nicola Marzari   Department of Materials Science and Engineering
13-5066   MIT   77 Massachusetts Avenue   Cambridge MA 02139-4307 USA
tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu 



More information about the Pw_forum mailing list