CCL:Summary: bencmarks on multiconfiguration methods

[Ping Lin <plin ^at^ mail.chem.tamu.edu>]

Dear All,
 Sorry for the late summary. The original question was:
 > I am looking for some benchmark information on different programs
 > running multiconfiguration methods. The methods are MCSCF(CASSCF), MRCI,
 > and CASPT2, etc. The programs I know include MOLCAS, MOLPRO, DALTON,
 > COLUMBUS, GAMESS-US, GAMESS-UK. I want to do some calculations with
 > these methods, but I do not know which one performs best.
 >
 Many people have expressed their interest on this subject, but not much
 information is available right now.
 I thank Dr. Dodi and Dr. N.O.J. Malcolm for the helpful discussion.
 There is no definite answer to this question from what I received.
 The benchmark here has two meanings:  the computational performances
 related to the results and to the CPU time. As far as I know, MOLPRO
 used internally contracted MRCI which can run much faster that other
 MRCI method . The CASPT2 have been used quite often for larger systems
 and seems to provide fairly good results . Form the level of theory, the
 results will be better and better from MCSCF(CASSCF) -> CASPT2 -> MRCI
 (internally contracted MRCI). For required CPU time, it will follow the
 reversed trend.
 That is all I can say now, I would appreciate any further information.
 --
 Ping Lin
 ------------------------------
 Department of Chemistry,
 Texas A&M University
 College Station, TX 77843-3255
 E-mail: plin ^at^ mail.chem.tamu.edu
 Tel: (409)845-8400  (Office)
      (409)862-9547  (Home)
 Fax: (409)845-4719