My Community

Dear All,
Recently, we installed vasp.5.2.11 to a new cluster, but came out several problems.

I tried to calculate a system, which contains 72 atoms/cell, but the total energy is converged to a positive value. I tested to the other systems, and the calculte values are quite reasonalbe. But, one system shows always positive value.

With a exact same input files, I've got a negative values when I run it on a previous cluster, but I cannot get a same value with a new cluster.

Any possible issues?

The previous version is vasp.5.2.11 18Jan11
And, a current one is vasp.5.2.11 25Nov10
Does this make any problems?

I checked "Free energy of the ion-electron system (eV)" in OUTCAR, and eigenvalues are positive and E-fermi is way to high (27.36).


Please give any comments.

Thanks,


<span class='smallblacktext'>[ Edited ]</span>

Hi,

check your libraries, e.g. fft, blas, etc ... if they run without complications on your new hardware/OS combination. You may have to update them to a newer version.

cheers,

alex