On-the-fly clustering for exascale molecular dynamics simulations. - presented by Dr Alizée Dubois and Thierry Carrard

On-the-fly clustering for exascale molecular dynamics simulations.

Alizée Dubois and Thierry Carrard

ADThierry Carrard
Slide at 36:57
Activités
Visionneur de documents
3 mars 15:38 0
pres_algo_killian.pdf
192.2%
TEST CASE : MICRO-JETTING - - 1 NODE CASE
MPIrank
CCId
Per-node MPI/OMP work
distribution :
4MPI X 32 OMP
60 106 atoms
CCvolume
14.106 voxels
A. DUBOIS - T. CARRARD - COMPUTER PHYSICS COMMUNICATIONS SEMINAR SERIES - 03/03/25
II Votre écran est partagé par le biais de l'applicationapp.zoom.us
Arrêter le partage
Masquer
Share slide
Summary (AI generated)

This slide presents a detailed overview of the test case utilized in our analysis. Similar to the previous scenario, we examined the algorithm developed by Killian, which represents a worst-case situation characterized by numerous connected components that are sparsely distributed across the domain. This algorithm demonstrates excellent scalability when handling many small connected components that are uniformly distributed throughout the simulation domain.

However, in this instance, we observe a significant connected component that extends widely. This phenomenon occurs due to the presence of numerous filaments that link groups of atoms together, resulting in a more complex structure within the simulation.