BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/New_York
X-LIC-LOCATION:America/New_York
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260422T143141Z
LOCATION:B314
DTSTART;TZID=America/New_York:20241117T160000
DTEND;TZID=America/New_York:20241117T163000
UID:submissions.supercomputing.org_SC24_sess741_ws_mlg104@linklings.com
SUMMARY:Scalable and Consistent Graph Neural Networks for Distributed Mesh
 -based Data-driven Modeling
DESCRIPTION:Shivam Barwey, Riccardo Balin, Bethany Lusch, Saumil Patel, Ra
 mesh Balakrishnan, and Pinaki Pal (Argonne National Laboratory (ANL)); Rom
 it Maulik (Pennsylvania State University, Argonne National Laboratory (ANL
 )); and Venkatram Vishwanath (Argonne National Laboratory (ANL))\n\nThis w
 ork develops a distributed graph neural network (GNN) methodology for mesh
 -based modeling applications using a consistent message passing layer. As 
 the name implies, the focus is on enabling scalable operations that satisf
 y physical consistency via halo nodes at sub-graph boundaries. Here, consi
 stency refers to the fact that a GNN trained and evaluated on one rank (on
 e large graph) is arithmetically equivalent to evaluations on multiple ran
 ks (a partitioned graph). This concept is demonstrated by interfacing GNNs
  with NekRS, a GPU-capable exascale CFD solver developed at Argonne Nation
 al Laboratory. It is shown how the NekRS mesh partitioning can be linked t
 o the distributed GNN training and inference routines, resulting in a scal
 able mesh-based data-driven modeling workflow. We study the impact of cons
 istency on the scalability of mesh-based GNNs, demonstrating efficient sca
 ling in consistent GNNs for up to O(1B) graph nodes on Frontier.\n\nTag: A
 rtificial Intelligence/Machine Learning, Graph Algorithms, Scalable Data M
 ining\n\nRegistration Category: Workshop Reg Pass\n\nSession Chairs: Seung
 -Hwan Lim (Oak Ridge National Laboratory (ORNL)); José Moreira (IBM); Cath
 erine Schuman (University of Tennessee, Knoxville); and Richard Vuduc (Geo
 rgia Institute of Technology)\n\n
END:VEVENT
END:VCALENDAR