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 24:21
tivités Visionneur de documents 3mars 15.25 pres_algo_killian.pdf 192,2% Q E STRONG SCALING RESULTS Analysis MD EAM 8.00 Ideal scaling 81.9M atoms Per-node MPI/OMP work 4.00 655.2M atoms distribution : 2.00 4 MPI X 32 OMP 1.00 0.50 CONSTANT WORKLOAD : 0.25 655.2 . 106 atoms 471.6 . 106 voxels 0.12 0.06 1024 2048 4096 Cores A. DUBOIS - T. CARRARD - COMPUTER PHYSICS COMMUNICATIONS SEMINAR SERIES - 03/03/25 II Votre écran est partagé par le blais de l'application app.zoom.us. Arrêter le partage Masquer
Share slide
Summary (AI generated)

In our analysis of weak scaling results, we maintain a constant load per core. As we increase the number of cores, we also increase the number of atoms proportionally. This approach allows us to evaluate the efficiency of our computation under varying workloads. The dotted black line represents the ideal scenario, and while we do not perfectly align with it, we observe that as the number of atoms increases, our performance improves relative to this ideal case.