TURBOMOLE Users Forum

Forum General => TURBOMOLE Forum General => Topic started by: brijesh on October 03, 2019, 09:20:53 AM

Title: Large system : too many cartesian basis functions
Post by: brijesh on October 03, 2019, 09:20:53 AM

Hi,

I am trying to run ridft for 370 atoms using Turbomole (huge) version. It gives me following error:

A HF calculation using the RI-J approximation will be carried out.
 Allocatable memory for RI due to $ricore (MB):         500
 Multipole approximation for Coulomb integrals is used
  nbfao =     11544 , ndi4 =     10000
 too many cartesian basis functions !

 ========================
  internal module stack:
 ------------------------
    molinp
 ========================

 fatal error in molinp.
 ridft ended abnormally
 ridft ended abnormally
program stopped.


Can you please help me?

Thanks.
Best Regards,
Brijesh

Title: Re: Large system : too many cartesian basis functions
Post by: uwe on October 03, 2019, 10:38:34 AM

Hi,

this seems to be a rather old version of Turbomole. Those limits have been completely removed many years ago.

Could you tell us, please, which Turbomole version you are using (is printed in the header of each output)?

Regards,

Uwe

Title: Re: Large system : too many cartesian basis functions
Post by: brijesh on October 03, 2019, 01:44:16 PM

Dear Dr. Uwe,

Here is the version:

ridft (b-cn0104.hpc2n.umu.se) : TURBOMOLE V6.4 2 Apr 2012 at 15:30:49
 Copyright (C) 2012 TURBOMOLE GmbH, Karlsruhe

I use:
define_huge
ridft -huge


Thank you.
Best Regards,
Brijesh