[Pw_forum] espresso 2.1.4 and ifc8.1.024

[Paolo Giannozzi]

On Thursday 07 July 2005 19:59, Patricia Paredes wrote:

> I made some experiments compiling all the versions of pwscf since
> 2.0.3 using ifc8.1.024, cluster mkl 7.1 and compiling mpi with the
> same compiler.  
>
> first iteration
>
> 2.0.3   117.11 secs
> 2.0.4   109.89 secs
> 2.1.1   110.03 secs
> 2.1.2   164.93 secs
> 2.1.3   152.52 secs
> 2.1.4   154.43 secs

this is what I got with your input - not sure about pseudopotentials -
for serial execution (2.4GHz dual Xeon, ifc7.1+mkl 7.01) of a single
SCF step:
Version CPU Time   Output File
  2.0.4 2h58m  http://web1.sns.it/~giannozz/pw/au.204
  2.1.4 2h54m  http://web1.sns.it/~giannozz/pw/au.214
  CVS  2h37m  http://web1.sns.it/~giannozz/pw/au.cvs
I don't see any overall performance degradation, even if the first 
scf iteration is actually slower than it used to be, due to different
requirements for the threshold of diagonalization

Paolo

-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy