dftd4 library does not support MPI parallelism

[cyxpku]

Details

The dftd4 library does not support running with multiple MPI processes (mpirun -np N with N > 1). ABACUS test cases involving dftd4 do not work with multiple MPI processes..

Affected ABACUS test cases:

• 03_NAO_multik / relax_cell_vdw4_d4
• 03_NAO_multik / relax_cell_vdw4_d4s

Question for dftd4 developers: Is MPI support planned or technically feasible in a future release?

Have you read FAQ on the online manual http://abacus.deepmodeling.com/en/latest/community/faq.html

• Yes, I have read the FAQ part on online manual.

Task list for Issue attackers (only for developers)

• Understand the problem or question described by the user.
• Check if the issue is a known problem or has been addressed in the documentation.
• Test the issue or problem on a similar system or environment, if possible.
• Identify the root cause or provide clarification on the user's question.
• Provide a step-by-step guide, including any necessary resources, to resolve the issue or answer the question.
• If the issue is related to documentation, update the documentation to prevent future confusion (optional).
• If the issue is related to code, consider implementing a fix or improvement (optional).
• Review and incorporate any relevant feedback from users or developers.
• Ensure the user's issue is resolved or their question is answered and close the ticket.

[Growl1234]

Question for dftd4 developers: Is MPI support planned or technically feasible in a future release?

No, there is no plan for an MPI support on the dftd4 program side; see dftd4/dftd4#329.

The only path to enable MPI parallelism for D4 correction in ABACUS has been made clear in #7380:

The only solution I can see would be to implement an ABACUS-native D4 backend, with an internal EEQ solver and MPI-parallel CN, pair, and three-body dispersion loops, while still relying on the DFT-D4 library for parameters, model data, and reference results.

There has been a successful implementation in CP2K; you can mainly see qs_dispersion_d4.F and eeq_method.F and use the master branch to test the effectiveness. It's clear that such implementation is not easy, especially for C++ coding.