[Wien] Wien compilation in IBM SP4

[Torsten Andersen]

Dear Alonso,

it would be nice to know under which conditions you got this result and 
why you think it is more wrong than the other result you refer to. 1mRy 
is not much... was it obtained with "run_lapw -ec 0.0001", or another 
convergence criterion?

Which SP4 configuration? CPU? Compiler? essl? operating system?

Best regards,
Torsten Andersen.

Roberto Alonso wrote:
> Dear Wien2k users
> 
> First of all, sorry if this information was communicated previously.
> I want to report that the compilation of the WIEN code with the -O4
> optimization in IBM SP4 gives wrong but not crazy results with the GGA
> xc. For example, total energies in fcc-Li differs with the correct value
> about 1 mRY. The problem is avoided compiling with the -O3 switch. LDA
> total energies are OK even with the -O4.
> 
> Best regards
> R.E.A.

-- 
Dr. Torsten Andersen                     TA-web: http://deep.at/myspace/
AG Hübner, Department of Physics, Kaiserslautern University, and
Condensed Matter Theory Group, Department of Physics, Uppsala University
Web: http://www.fysik4.fysik.uu.se/         http://www.physik.uni-kl.de/