Recent Posts

Pages: [1] 2 3 ... 10
1
ccsdf12 / Re: (T) bit of CCSD(T) calculation taking forever
« Last post by uwe on Today at 11:51:25 am »
Hi,

yes, MKL_ENABLE_INSTRUCTIONS is one option. Another option is to change the allowed memory for the intermediate steps ($maxcor). This changes the number of MKL calls and also the size of the individual steps and could help too... but that's a bit of a lottery.

The TM support team is still there and not lost, so do hesitate to get in touch with them.
2
ccsdf12 / Re: (T) bit of CCSD(T) calculation taking forever
« Last post by martijn on Today at 11:36:57 am »
Thanks Uwe! Would that involve setting: export MKL_ENABLE_INSTRUCTIONS=SSE4.2 ? If so I'll have a look.

In the meantime, I've successfully finished the calculation with version 7.01 (same as with the Raman script issue a while back), suggesting indeed that this is a problem with some sort of code optimisation in later versions rather than an inherent bug.

I'll raise the issue with TM support when I've figured out how you do that nowadays, now cosmologic has become part of Dassault.
3
ccsdf12 / Re: (T) bit of CCSD(T) calculation taking forever
« Last post by uwe on May 28, 2020, 09:39:50 am »
Hello,

the triples part should in general take approximately as long as the doubles (the time from start till the final CCSD energy is printed).

The (T) part heavily uses linear algebra routines, which are done by using Intel's multi-threaded library (MKL). There have been cases where ccsdf12 tried to generate a new thread and waited for it to be started for an extremely long time - that was observed on CPUs with AVX2 or AVX512 only.

You probably did run into this problem, but there are work-arounds for it. Telling MKL not to use AVX2 does usually help, but it will run a factor of two slower (on CPUs where this does not happen). So I'd be careful to use this approach in general.

Could you please report that to the Turbomole Support? It is also important to make sure that the next Turbomole release will not run into this specific error.

Regards,
Uwe
4
ccsdf12 / (T) bit of CCSD(T) calculation taking forever
« Last post by martijn on May 26, 2020, 06:52:04 pm »
Hi,

I have been trying to run a CCSD(T)/def2-TZVPP using Turbomole 7.41 on a coordination complex containing one Ni atom, eight heavy atoms and for hydrogens (412 basis functions). The CCSD bit of the calculations runs fine, taking 7 hours on 8 of the not newest cores but now 4 days later I have seen no progress on the (T) bit of the calculation. The last thing in the output file of the still running calculation is the end of the CCSD bit:

     **********************************************************************
     *                                                                    *
     *   RHF  energy                             :  -3250.7809131050      *
     *   correlation energy                      :     -2.9299415667      *
     *                                                                    *
     *   Final CCSD energy                       :  -3253.7108546717      *
     *                                                                    *
     *   D1 diagnostic                           :      0.3059            *
     *                                                                    *
     **********************************************************************


and non of the temporary files in the directory have changed since this was written to the output, even if according to top the ccsdf12_omp process is happily beavering away.

Is the calculation of the pertubative triples really others of magnitude more expensive than the CCSD part or has something likely gone wrong?

Thanks,

Martijn
5
Statpt / TS searching - problem with hessian (macroiterations)
« Last post by yasmeen on May 26, 2020, 10:42:31 am »
During TS searching, to improve procedure  I put hssfreq 3 in $statpt. To make it less time-consuming I added in control $les all 1. But in some cases it failed because 25 macro-iterations are not enogh to converge. I changed it to 35 but it didn't help neither.
What is possible reason of it and what could be a solution?
I haven't faced with such a problem before.
Thank you for any help.
6
Dear Uwe,

Thanks a lot!

I have tried full Hessian calculated every N'th step by hssfreq 1. It is indeed very time consuming. I checked the vibration modes so far. It is still good. By that I mean I can see the vibration along the reaction coordinates.

I will see if it works out. Otherwise, I will try to use keeptmode option in the $statpt section.

Best regards,
Dr. Shaotao Bai
KU Leuven
7
Hello,

the transition state search by default follows the mode which is closest to the initial transition vector. The statpt module calculates the overlap of the initial mode with those of each step and then follows the one which has the largest overlap.

First thing to check before running a transition state search after a frequency calculation is to make sure not having any approximate hessian files in your job directory - only the newly generated hessian file. So remove the file hessapprox if it is present in the job directory.

To always follow the mode with the lowest eigenvalue, try the keeptmode option in the $statpt section (see manual for details, search for keeptmode).

If the two lowest modes come get very close to each other, such that the overlaps of the initial eigenvector with both modes are similar, it might help to run a full Hessian calculation from time to time. An automated way can be activated by adding the hssfreq <number> keyword to $statpt, and the full Hessian will be calculated every N'th step. This can be very time consuming, though.

Good luck!

Uwe
8
Dear friends,

I have found one structure that should close to the transition state structure in woelfling-job calculations. Aoforce calculations show two negative frequencies as shown below. The mode 1 with frequency of -527.58 cm -1 is the imaginary frequency along the reaction path. I have setted itrvec  1 for transition state search, which leads to the structure that actually follows mode 2. Thus, the problem is what number should I set for this job? Another general question: how can I know I follow the right transition vector for transition state search?

$vibrational spectrum
#  mode     symmetry     wave number   IR intensity    selection rules
#                         cm**(-1)        km/mol         IR     RAMAN
     1        a            -527.58         0.00000       YES     YES
     2        a             -78.66         0.00000       YES     YES
     3                       -0.00         0.00000        -       -
     4                       -0.00         0.00000        -       -
     5                       -0.00         0.00000        -       -
     6                       -0.00         0.00000        -       -
     7                        0.00         0.00000        -       -
     8                        0.00         0.00000        -       -
     9        a              12.93         0.00928       YES     YES
    10        a              25.73         0.00596       YES     YES
    11        a              31.46         0.26516       YES     YES
    12        a              35.84         0.23318       YES     YES
    13        a              44.13         0.02108       YES     YES
    14        a              45.94         0.08892       YES     YES
Many thanks in advance!

Best regards,
Dr. Shaotao Bai
KU Leuven
9
Aoforce and Numforce / reduced mass of diatomic molecule in aoforce output
« Last post by tonyst on May 11, 2020, 02:04:30 pm »
Dear Turbomole enthusiasts,

an analytical force constant calculation of IrN yielded the following output:

Code: [Select]
       mode               1        2        3        4        5        6

     frequency           0.00     0.00     0.00     0.00     0.00  1239.80

     symmetry                                                         a1 

        IR                -        -        -        -        -       YES
|dDIP/dQ|   (a.u.)     0.0000   0.0000   0.0000   0.0000   0.0000   0.0056
intensity (km/mol)       0.00     0.00     0.00     0.00     0.00    56.16
intensity (  %   )       0.00     0.00     0.00     0.00     0.00   100.00
 
       RAMAN              -        -        -        -        -       YES
 
  1   ir          x   0.00000  0.00000  1.00000  0.00000  0.00000  0.00000
                  y   0.00000  0.00000  0.00000  0.00000  1.00000  0.00000
                  z   0.00000  0.70711  0.00000  0.00000  0.00000 -0.07268
  2   n           x   0.00000  0.00000  0.00000  1.00000  0.00000  0.00000
                  y   1.00000  0.00000  0.00000  0.00000  0.00000  0.00000
                  z   0.00000  0.70711  0.00000  0.00000  0.00000  0.99736

reduced mass(g/mol)    14.007  103.113  192.220   14.007  192.220   14.948

The reduced mass is calculated to be 14.948, but should it not be 192.220*14.007/(192.220+14.007)=13.056? What am I missing?

Thank you for your help

/Edit: and for N2 it is 14?

Code: [Select]
       mode               1        2        3        4        5        6

     frequency           0.00     0.00     0.00     0.00     0.00  2348.41

     symmetry                                                         a1g

        IR                -        -        -        -        -       NO
|dDIP/dQ|   (a.u.)     0.0000   0.0000   0.0000   0.0000   0.0000   0.0000
intensity (km/mol)       0.00     0.00     0.00     0.00     0.00     0.00
intensity (  %   )       0.00     0.00     0.00     0.00     0.00     0.00
 
       RAMAN              -        -        -        -        -       YES
 
  1   n           x   0.00000  0.00000  0.00000  1.00000  0.00000  0.00000
                  y   0.00000  1.00000  0.00000  0.00000  0.00000  0.00000
                  z   0.70711  0.00000  0.00000  0.00000  0.00000  0.70711
  2   n           x   0.00000  0.00000  1.00000  0.00000  0.00000  0.00000
                  y   0.00000  0.00000  0.00000  0.00000  1.00000  0.00000
                  z   0.70711  0.00000  0.00000  0.00000  0.00000 -0.70711

reduced mass(g/mol)    14.003   14.003   14.003   14.003   14.003   14.003

/Edit2:

Taking the wavenumber and force constant from that calculation I derive a reduced mass of 13.055474:

Code: [Select]
You have: (0.7594293 hartree / atomiclength^2)/(1239.80 cm^-1 * c * 2 * pi)^2*N_A
You want: g/mol
* 13.055474

I am really curious what that reduced mass in the aoforce output means.

10
Aoforce and Numforce / Re: Visualization vibration modes? aoforce2g98
« Last post by baishaotao on May 07, 2020, 12:46:27 pm »
Hello Uwe,

Many thanks. It works now.

Best,
Dr. Shaotao Bai
KU Leuven
Pages: [1] 2 3 ... 10