Hello all,
On today’s meeting, I would like to start more concrete discussions on re-implementing tabulated interaction support.
This functionality is important for our group. We would like to finally be able to bring the SPICA coarse grain forcefield to a modern version of GROMACS with GPU acceleration.
With the quarterly planning meeting fast approaching, my aim would be to discuss what changes would need to be made, and features (re-) implemented. A couple of months ago, I started a similar discussion and was told in broad strokes what would need to happen (e.g. implementing support for multiple pair lists broken down by interaction type). My hope is to gather more concrete opinions and ideas from the core developers. This will help me prepare a roadmap to present at the planning meeting and to create a proper issue to track this proposal.
Specific points to address (off the top of my head):
- Are there any (and if so, what specifically) blockers for this feature?
- My current understanding is that this would require the re-implementation of the group cut-off scheme. Is this something desireable? Or should I look for a different approach?