We have recently developed a lattice-free simulation program in computational cell biology which needs the introduction and management of the biomechanical interactions of cells. These interactions are associated with short range forces which act on nearest-neighbors only. The forces act in the rearrangement of cells due to proliferation and cell growth and this requires a recalculation of the proximity relations at each time step. Here we describe the implementation of an algorithm to efficiently compute the proximity relations and designed to run on Graphics Processing Units (GPUs). The results of the first test runs on an NVidia Fermi GPU are encouraging: the algorithm has the potential to significantly boost the simulation program and to map the disordered lattice also on other multicore machines with hypercubic connectivity.
Neighbor search algorithm for lattice-free simulations with short-range forces
CHIGNOLA, Roberto
2014-01-01
Abstract
We have recently developed a lattice-free simulation program in computational cell biology which needs the introduction and management of the biomechanical interactions of cells. These interactions are associated with short range forces which act on nearest-neighbors only. The forces act in the rearrangement of cells due to proliferation and cell growth and this requires a recalculation of the proximity relations at each time step. Here we describe the implementation of an algorithm to efficiently compute the proximity relations and designed to run on Graphics Processing Units (GPUs). The results of the first test runs on an NVidia Fermi GPU are encouraging: the algorithm has the potential to significantly boost the simulation program and to map the disordered lattice also on other multicore machines with hypercubic connectivity.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
