CCL:G: Package for CCSD(T)
- From: "Kalju Kahn" <kalju..chem.ucsb.edu>
- Subject: CCL:G: Package for CCSD(T)
- Date: Mon, 20 Jul 2009 12:06:52 -0700
Sent to CCL by: "Kalju Kahn" [kalju()chem.ucsb.edu]
Neha,
I have found DIRCCR12 a good free choice on a desktop system. Water in
aug-cc-pV6Z basis (443 bf, run with very tight SCF, integral, and CC
convergence criteria) took 754 min on dual-core E8400 ~ 3.00GHz with 8 GB
RAM
Please note that you need to tighten some convergence criteria with aug
basis (how much depends on your needed accuracy, of course). And when you
have more than about 500 basis functions, you will need more than 8 GB of
RAM.
You can get the software (after e-mailing Dr. Jozef Noga) from
http://www-laog.obs.ujf-grenoble.fr/~valiron/ccr12/.
Psi3 (http://www.psicode.org/index.html) is another nice choice if
you
need a very fast and free code. However, I spent a significantly more
effort getting it working (at least with the old 3.2 version), and
experienced a share of crashes during some more demanding calculations.
Formic acid CCSD/cc-pV5Z calculation (518 AO, 383 symmetry orbitals in Cs
symmetry) completed in 1100 minutes with default settings on (now almost
too old) Core 2 6300 ~ 1.86GHz.
I think most CC codes will use disk for reading and writing very large
files, so make sure that you have an appropriate hardware (minimally a
pair of fast disks in RAID0 for scratch space) to support these
calculations.
Best wishes,
Kalju
>
> Sent to CCL by: "Neha Awasthi" [neha.awasthi-#-gmail.com]
> Hello CCL community,
>
> I wanted to ask for advise for a reliable program for performing CCSD and
> CCSD(T) calculations using (aug)-cc-pVxZ (where x = D, T, Q...) basis
> sets.
> I have tried Gaussian, G03, and it turns out to be very inefficient (save
> large scratch files, and slow) for small molecules upto ~5 atoms.
>
> Please let me know which package/program is considered reliable as well as
> has good performance.
>
> It would be ideal to have one with a free license or free to download :)
> I will compile all responses and post them.
>
> Thank you,
> Neha>
>
>
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Dr. Kalju Kahn
Department of Chemistry and Biochemistry
UC Santa Barbara, CA 93106