Is it possible to overcome the restriction to 32 active orbitals?

Peter Knowles P.J.Knowles at bham.ac.uk
Thu Feb 14 13:29:43 GMT 2002


Sorry, but it's absolutely not possible to exceed 32 active orbitals
because bit patterns that occupy two bits per active orbitals are used
throughout the code. I would anyway be somewhat sceptical that one could
really do a contracted MRCI computation with so many active orbitals
on any available hardware, unless one is talking about just 2 or 3
electrons, for which other methods might provide a better solution.

At Thu, 14 Feb 2002 13:10:38 +0100,
Philipp Spuhler wrote:
> 
> [1  <text/plain; us-ascii (7bit)>]
> Hello,
> 
> I tried to run a MRCI calculation with 42 active orbitals on a machine
> which allows to use 64 bit integers, but the limit is obviously 32
> active orbitals. I would like to know if it is possible to overcome this
> restriction by configuring the program appropriately during the
> compilation or if this limit is given by  the 64bit architecture of the
> hardware and therefore cannot be exceeded. (According to the user's
> manual there is a restriction on 32 bit machines which is 16 active
> orbitals. So it seems to be a similar situation!?)
> 
> Does anybody have an idea??
> 

--
Prof. Peter J. Knowles              
Email P.J.Knowles at bham.ac.uk  Phone +44-121-414-7472  Fax +44-121-414-7471
School of Chemical Sciences, University of Birmingham, Birmingham, B15 2TT, UK
WWW http://www.tc.bham.ac.uk/~peterk/



More information about the Molpro-user mailing list