[Pw_forum] the relax is to slow

vega vegalew at hotmail.com
Fri Sep 12 06:14:47 CEST 2008


pay more attention on you CPU user% ratio and see whether your cpu is waiting for IO or virtual memory.
if IO wait% quite high, this may be because of the network. as far as I know, the network will quite busy if
you use very few k points. you may need infiniband for gamma point calculation for a large system.  

If your memory is not enough, this will also happen. More atoms in the system will demand much more memory.
 Virtual memory will be employed if the physical memory is not enough. And the exchanging of  data between 
the virtual memory and physical memory need a lot of time. So you may suffer from this.

I know nothing about your hardware, and input file. so I can't tell you the further about your problem.

hope helps

vega 


=================================================================================
Vega Lew (weijia liu)
PH.D Candidate in Chemical Engineering
State Key Laboratory of Materials-oriented Chemical Engineering
College of Chemistry and Chemical Engineering
Nanjing University of Technology, 210009, Nanjing, Jiangsu, China


From: wangqj1 
Sent: Friday, September 12, 2008 9:33 AM
To: pw_forum 
Subject: [Pw_forum] the relax is to slow



 Dear pwscf users 
      When I use 'relax' to optimize the suppercell of 48 atoms ,it runs too slow ,in 9 hours the iterative is only 5,I think it will need 1 month to relax the suppercell like this,but I remember several months ago ,I relax the suppercell of 32 atoms ,it only needs 8 atoms .I don't know what's the reason cause this ,how can I to promote this .
      Any advice will be appreciated !
sincerely
qjwang
XiangTan University
       
      




--------------------------------------------------------------------------------


_______________________________________________
Pw_forum mailing list
Pw_forum at pwscf.org
http://www.democritos.it/mailman/listinfo/pw_forum
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080912/5c488ea6/attachment.htm 


More information about the Pw_forum mailing list