[CPMD-list] Free Energy on Cray and IBMSp4

[DR.ALDO ROMERO]

I am trying to do a calculation of a metallic system by using the Free
energy functional, I have initialize the wavefucntion randomly.
When I run the calculation on the Cray T3E I get the errors:

  FRIESNER| VERY BAD EIGENVALUE !!    

but when I run the same calculation on the IBM Sp4, the calculation
runs cleanly. Is there an specific problem on the Cray?. Any other
calculation by using the normal functional such as geometry
optimization or wavefunction optimization runs fine.

Thanks in advance.

-------------------------------------------------------------------------
Prof. Aldo Humberto Romero
IPICyT
Camino a la presa San Jose 2055
Colonia Lomas 4a seccion
C.P 78216
San Luis Potosi, SLP
Mexico
 
email:  aldo at ipicyt.edu.mx
Phone:  (52)-444-8342000 Ext 2038
Fax:    (52)-444-8342010
http://materials.ipicyt.edu.mx/