Issue with running the Si–VASP example: empty Results files and limited CalcFold generation

[Sparsh Sharma]

Dear USPEX community,

We are trying to run the Si example provided for VASP, but we are encountering some issues during the calculation.

Specifically, the final enthalpies and individual structure files are not being generated properly in the results folder. Instead, the corresponding files are being created as empty files.

The VASP OUTCAR files, however, are being generated successfully. We also noticed that the number of CalcFold directories generated appears to depend on the numParallelCalcs parameter. For example, if we set:

numParallelCalcs = 3

only 3 CalcFold directories are generated, whereas setting:

numParallelCalcs = 10

generates 10 CalcFold directories.

Our understanding is that the number of structures/calculations generated should ultimately correspond to the specified initial population size, with numParallelCalcs controlling the number of calculations running in parallel rather than limiting the total number of structures evaluated.

We are attaching our INPUT.txt and batch submission file below for reference.

Could someone please help us identify what we might be doing incorrectly and advise us on how to correct the setup?

Thank you very much for your help.

[Ishita Das]

Hello everyone,

Just following up on this question in case it was missed. We are still facing the same issue with the VASP Si example in USPEX, particularly the empty files in the results folder and the behavior of numParallelCalcs.

We would really appreciate any guidance on whether this is a configuration/submission issue or if there is something specific we should check in the USPEX–VASP setup.

Thank you!