RudolfWeeber
RudolfWeeber
Due to their propagation at different points in the integration loop before and after force calculation the state of the virtual sites and normal particles are inconsistent. It needs to...
* Removs the less-than-transparent shifting logic which prevents posiiton folding of vs relative. this is no longer needed, since we now use minimum_image_distance everywhere. * Apply Lees Edwards position shift...
Description of changes: this is an early sketch for a cell system based on RegularDecomposition, but for simulations on a singel mpi rank. It does not have ghost cells and...
Currently, Es calculates forces/energies in the RegularDecomposition as follows: \ * Iterate over all local (not ghost) cells * For each of them consider all upper/right/back neighbor cells (i.e. some...
LB ignores Lees Edwares boundary condition, if it is set up after LB. To reproduce, switch the initializaiton order from 1. LE 2. LB to the opposite in the Lb...
contains #4827 This * fixs index calculation in the LB Walberla sweep doing the interpolation of populations at the Lees-Edwards boundary * adds missing Lees-Edwards velocity shift in the lb...
My current understanding: * Currently it seems best to focus at the Lennard-Jones benchmark. * Lennard-Jones (together with other short range forces) are most relevant on the particle side *...
There seem to be a number of performance regression. Between 4.0 and the current python branch, the lj test deteriorates from 0.41ms to 0.6ms on gerenuk. I tried to keep...
In preparation for performance portable electrostatics. ## ESPResSo interaciton * Theere is still some immedaite dependent on ESPResSo data structures in the P3M code, e.g., * the System class *...