[Pw_forum] pwscf tests/examples question

[mbaris at metu.edu.tr]

Hello to all,
   First of all, I must mention that I am completely a newbie, and some 
of my questions may be too trivial, but I will be glad if you answer 
them.

    I recently compiled espresso 3.2.2 on an Intel box (details at the 
end), the compilation went smoothly, but I am encountering some errors 
in tests/examples.

First things first, when I edit the environment_variables file to use 
mpich2 (PARA_PREFIX="mpiexec -n 4") and run run_all_examples, the batch 
halts at   example05 in running pp.x. On a single cpu (i.e. with 
PARA_PREFIX empty) upto example34 can be run, which fails stating vdw.x 
can not be found. Is this a bug, or something related to my computer? 
(I failed to see the error in example05 script)

Secondly, (and this is the most important problem I encounter)
In some of the examples, there are deviations from the reference data, 
most of them are not worth mentioning, as they are not in the final 
structure, however there is one puzzling me and my colleague greatly. 
Looking at c.scf.out in example02 (>:reference <:my computer) :
<      estimated scf accuracy    <        4.3E-10 Ry
---
>      estimated scf accuracy    <        4.4E-10 Ry
207,209c193,195
<      one-electron contribution =     7.88010983 Ry
<      hartree contribution      =     1.91301246 Ry
<      xc contribution           =    -7.07049550 Ry
---
>      one-electron contribution =     7.88010934 Ry
>      hartree contribution      =     1.91301318 Ry
>      xc contribution           =    -7.07049573 Ry
As you can see, the contribution weights are different, however the 
resulting energy is the same!!! check_example seems to ignore this. I 
have tried running the example using version 3.2 and 3.2.1 to no avail, 
always the same. What might be the problem, or is it a problem I need 
to dig further?

Lastly, with the encouragement of the manual I have installed intel 
MKL, but instead of seeing a performance leap, the code runs slightly 
slower now (order of 0.01 sec). I guess more fine tuning is required on 
my side.

This box is a test bed for a highly parallel blade server array in my 
school, I am using it to prepare mpich2 integrated gridengine scripts 
and such (which I am happy to share if demanded) along with the 
projects I am assigned.

Thank you for your time,
O. Baris Malcioglu  METU Physics Dep. Ankara

Intel box details:
Hardware:
  Intel D975xbx2
  Intel core 2 quad 6600
  2 gb ddr2 667 4-4-4-15 ram
  Raid 5 sata disk array.
Software:
  OpenSuse 10.2 64 bit (Stock Kernel 2.6.18.8-0.3, stock glibc 2.5-25)
  Intel fortran compiler 64bit 9.1.045
  Intel c compiler 64 bit 9.1.049
  Intel mkl 9.0
  mpich2 with smpd daemon (for tight integration with sge)
  Sun grid engine 6.1