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 14:19
METHODOLOGY AND TECHNICAL BOTTLENECKS
Alizee Dubois and Thi.
CONNECTED COMPONENTS ANALYSIS
IMAGE PROCESSING
GRAPH PROCESSING
COMMUNITY
COMMUNITY
Shared Memory Parallelization
Distributed Memory
Finite number of pass
Iterative algorithm
Not scalable beyond a single
Scalable, but requires multiple
node
iterations
Necessity to port the algorithm
To be adapted for 3D images
to a distributed memory system
while limiting the number of
passes
A. DUBOIS - T. CARRARD - COMPUTER PHYSICS COMMUNICATIONS SEMINAR SERIES - 03/03/25
II Votre écran est partagé par le biais de l'application app.zoom.us.
Arrêter le partage
Masquer
Share slide
Summary (AI generated)

The analysis begins by subdividing the space into cells, which allows us to define subcells or box cells. Each atomic quantity is projected onto these boxes, resulting in a 3D voxelized image. Our initial approach will be to implement the image processing methodology developed by the COMMUNITY, as it aligns closely with our existing data format.

This work is based on the PhD thesis of Kan, who conducted the foundational research. I am here to present his findings and provide a comprehensive overview of the work he has accomplished.