Some times ago I posted a question:
"Does anybody have gained experience with the Jaguar and/or Titan
program?
Do they calculate molecular geometries and frequencies substantially
faster than GAUSSIAN 94/98?"
Here is the summary of the answers. Thanks to all who replied.
--------------------------------------------------------------
I've been using Jaguar for DFT calculations on transition metal complexes
for about a year now and have
been very happy with it. In my hands I've found much faster than G94
for both geometry optimization and
frequency calculations. I have to note that for the systems I've been
working with the frequency calculations
are done numerically instead of analytically but for my purposes they're
fine. Schrodinger's tech support
are also quite helpfull and fast.
--------------------------------------------------------------
I've used jaguar a fair bit, as well as gaussian 98, and have found
jaguar to be significantly faster for the same systems at the same
level
of theory. This is on PC's runnning linux, i don't have experience
with
windows, which from you're query about titan, i suppose you are using.
--------------------------------------------------------------
Several years ago I ran a few B3LYP comparisons (single point, no
optimizations)
between Jaguar and G94
on small molecules. Jaguar was much faster. My guess is that this speed
difference would be maintained on
larger molecules, but I did not do the test.
--------------------------------------------------------------
I've found Jaguar to be 7 - 10 times faster than G98 for geometry
optimisations and frequency calculations of organometallic compounds.
Jobs
were run on Linux/Intel or Dec Alphas.
--------------------------------------------------------------
Bye
Th. Mehnert