Population size of Crystal Structure Prediction Scripter

[Roc2019]

Dear Sir,
It is aviliable  to run 5 work (DFT or NEGF) simultaneously according to my license. 
However, when I did a test about the Crystal Structure Prediction Scripter such as the example of (TiO2)2,  it will print error due to license number if the parameters Population size and Number of elites are larger than 5.
In addition, if these two parameters are set to 5 or smaller, it cannot get the convergened result even when the Number of generations is set to 100000.

So, how to solve this problem?

Thank you.

Roc

[Anders Blom]

First try to limit the number of MPI processes to 5 and see if it runs. After that, you can increase to 10, 20 etc if you also use set a ParallelParameters on the calculator, where you can specify processes_per_individual to be 2, 4, etc to make it go faster. Let us know if this works as intended.

[Roc2019]

Thank you so much, Dr. Blom.

I did some tests according to your suggestions.  This method seems not very effective.  When the number of MPI processes is set to 5, Population size and Number of elites are set to 10, it sometimes works and sometimes does not work due to the license number. 

In addition,  is it possible to search the structure of two-dimensitional structure like graphene, not only the bulk structure?   If yes, how to set the input file?

Thanks again.

Roc