[Pw_forum] About the difference between the energy obtained from geometry optimization and single energy calculation.
Dear PWSCF users, The geometry optimization can give the minimal energy of our system. What's the difference between the energy obtained from geometry optimization and the single energy calculation? Best regards. -- Hongsheng Zhao School of Physics and Electrical Information Science, Ningxia University, Yinchuan 750021, China GnuPG DSA: 0xD108493 2010-12-24 __ ? http://cn.mail.yahoo.com
[Pw_forum] Merry Christmas
Mearry Christmas to every body. Good luck for us!
[Pw_forum] ion_radius
Dear Forum, I am using espresso-4.2.1. In example18, for cp.x practice. I found the " ion_radius(1) and ion_radius(2) " in the input file. Could you please tell me something about this input parameter, and how to set? as following: ion_dynamics='verlet' , ion_temperature='nose' , tempw = .2 , fnosep = 250 , ion_radius(1) = 1.0 , ion_radius(2) = 1.0 , I can not find any explanation about ion_radius from espresso manual. Thanks very much. WANG Riping 2010.12.24 -- ** WANG Riping Ph.D student, Institute for Study of the Earth's Interior,Okayama University, 827 Yamada, Misasa, Tottori-ken 682-0193, Japan Tel: +81-858-43-3739(Office), 1215(Inst) E-mail: wang.riping.81 at gmail.com ** -- next part -- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20101224/f278e2a2/attachment.htm
[Pw_forum] how does the calculation scales with number of processors
Thanks a lot Dr. Stefano for your reply. On Fri, Dec 24, 2010 at 3:50 PM, Stefano de Gironcoli wrote: > It strongly depends on the particular calculation and on the > parallelization strategy you choose. Linear scaling would be the ideal > scaling which you never get. > The things to keep in mind are load balancing and communication > overhead. This last is affected by your communication network latency > and bandwidth ... As a general rule the bigger the calculation the > easier is to scale. > It may be useful to have a look at the final timing summary of a > calculation to see the fraction of time spent in communication (FFT > scatter/gather and reduce operations) compared with the total wall clock > time for your system and how this changes for different settings. > Another factor to keep in mind is RAM memory as with certain > parallelization strategies you can trade some speed with increased memory. > > stefano > > mohnish pandey wrote: > > Dear QE users, > > I am trying to see how the runtime scales with > > number of processors. I did the same calculation using one node with > eight > > processors on one cluster and 8 nodes with total 56 processors on other > > cluster but the time does not seem to scale linearly with the number of > > processors. Can anybody give me an idea how does the time scale with > number > > of processors. > > Thanks a lot in advance. > > > > > > > > > > ___ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > ___ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Regards, MOHNISH, - Mohnish Pandey Y6927262,5th Year dual degree student, Department of Chemical Engineering, IIT KANPUR, UP, INDIA ----- -- next part -- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20101224/e6654593/attachment.htm
[Pw_forum] how does the calculation scales with number of processors
Dear QE users, I am trying to see how the runtime scales with number of processors. I did the same calculation using one node with eight processors on one cluster and 8 nodes with total 56 processors on other cluster but the time does not seem to scale linearly with the number of processors. Can anybody give me an idea how does the time scale with number of processors. Thanks a lot in advance. -- Regards, MOHNISH, - Mohnish Pandey Y6927262,5th Year dual degree student, Department of Chemical Engineering, IIT KANPUR, UP, INDIA - -- next part -- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20101224/094c0e6d/attachment.htm
[Pw_forum] how does the calculation scales with number of processors
On Dec 24, 2010, at 10:42 , mohnish pandey wrote: > Can anybody give me an idea how does the time scale with number of > processors. please see the user guide in particular chapter 3: http://www.quantum-espresso.org/user_guide/node15.html and this, in particular the "(Too) Frequently Asked Question" http://www.fisica.uniud.it/~giannozz/QE-Tutorial/ P. --- Paolo Giannozzi, Dept of Chemistry, Univ. Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222
[Pw_forum] how does the calculation scales with number of processors
It strongly depends on the particular calculation and on the parallelization strategy you choose. Linear scaling would be the ideal scaling which you never get. The things to keep in mind are load balancing and communication overhead. This last is affected by your communication network latency and bandwidth ... As a general rule the bigger the calculation the easier is to scale. It may be useful to have a look at the final timing summary of a calculation to see the fraction of time spent in communication (FFT scatter/gather and reduce operations) compared with the total wall clock time for your system and how this changes for different settings. Another factor to keep in mind is RAM memory as with certain parallelization strategies you can trade some speed with increased memory. stefano mohnish pandey wrote: > Dear QE users, > I am trying to see how the runtime scales with > number of processors. I did the same calculation using one node with eight > processors on one cluster and 8 nodes with total 56 processors on other > cluster but the time does not seem to scale linearly with the number of > processors. Can anybody give me an idea how does the time scale with number > of processors. > Thanks a lot in advance. > > > > > ___ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum >
[Pw_forum] ion_radius
On Dec 24, 2010, at 7:56 , Riping WANG wrote: > In example18, for cp.x practice. I found the " ion_radius(1) > and ion_radius(2) " in the input file. > Could you please tell me something about this input parameter, and > how to set? usuallly you do not need to set it: the default value should be fine. ion_radius is the width of a gaussian distribution of charge replacing the point charge of the (pseudo-)nucleus. It should be big enough to ensure convergence of the total energy in G-space, small enough to yield negligible overlap between gaussians. Valus around 0.8 are usually fine. The total energy and forces should not change if ion_radius changes. P. --- Paolo Giannozzi, Dept of Chemistry, Univ. Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222
