How to deal with excessive memory usage

[bubble]

Dear,
Recently, I try to test a IV curve of a device configuration. The calculation is always abnormally terminated because of  "excessive memory usage". I decrease the basis sets (from DZP to SZP) and K point (from 100 to 20) but the problem still exists. The attachment is the input file.

| Real space grid sampling is (219, 271, 427) in a, b, and c directions.       |
|                                                                              |
+------------------------------------------------------------------------------+
+------------------------------------------------------------------------------+
|                                                                              |
| Storage of atomic core grids requires   193 MB per grid                      |
|                                                                              |
+--------------------------------------------------------------------------


The above information is obtained from the log file. My comprehension is that the total memory needed is 219X271X427X193 MB=4776377 GB, a very large number and can not achieve in any super computor . I wonder why such a simple system (only contains C and H atoms) needs so many memory?

Thank you in advance.

[Anders Blom]

No, you misread it. It will just need 193 MB in total. Maybe you run out of memory because you run too many MPI processes on the same machine? You can check the total memory that the calculation will need by clicking the button in the Script Generator.

[bubble]

Thank you for your reply, Blom,
     During the calculation, according to your guide, I used 32 cores with 16 MPI. However, the problem still existed. In addition, the total memeory advised by ATK calculater is about 2000 MB (=2 GB).
    Could you give me some further suggestions?

[Anders Blom]

2 GB per MPI process yes. So it's 32 GB + perhaps a bit (the memory estimate is not very accurate) for 16 MPIs.