GROMACS version: 2023
GROMACS modification: No
Currently, there are two possible implementation for QMMM, one is through CP2K and the other is through IBM CPMD, performance and ease of use, which would you advice to use?
I think for CP2K, the GEEP feature for QMMM is still being planned for GPU. On the other hand, CPMD source code seems to indicate the presence of GPU implementation via CUDA, however, there is a lack of documentation up to what extent is it implemented, also, its new in town for the open source.