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:B310
DTSTART;TZID=America/New_York:20241117T120500
DTEND;TZID=America/New_York:20241117T123000
UID:submissions.supercomputing.org_SC24_sess737_ws_ia119@linklings.com
SUMMARY:NEO-DNND: Communication-Optimized Distributed Nearest Neighbor Gra
 ph Construction
DESCRIPTION:Keita Iwabuchi, Trevor Steil, Benjamin Priest, Roger Pearce, a
 nd Geoffrey Sanders (Lawrence Livermore National Laboratory (LLNL))\n\nGra
 ph-based approximate nearest neighbor algorithms have shown high neighbor 
 structure representation quality.\nNN-Descent is a widely known graph-base
 d approximate nearest neighbor (ANN) algorithm.\nHowever, graph-based appr
 oaches are memory- and time-consuming.\n\nTo address the drawbacks, we dev
 elop a scalable distributed NN-Descent.\nOur NEO-DNND (neighbor-checking e
 fficiency optimized distributed NN-Descent) is built on top of MPI and des
 igned to utilize network bandwidth efficiently.\nNEO-DNND reduces duplicat
 e elements, increases intra-node data sharing, and leverages available DRA
 M to replicate data that may be sent frequently.\n\nNEO-DNND showed remark
 able scalability up to 256 nodes and was able to construct neighborhood gr
 aphs from billion-scale datasets.\nCompared to a leading shared-memory ANN
  library, NEO-DNND achieved competitive performance even on a single node 
 and exhibited 41.7X better performance by scaling up to 32 nodes.\nFurther
 more, NEO-DNND outperformed a state-of-the-art distributed NN-Descent impl
 ementation, achieving up to a 6.0X speedup.\n\nTag: Graph Algorithms, Hete
 rogeneous Computing, Programming Frameworks and System Software\n\nRegistr
 ation Category: Workshop Reg Pass\n\nSession Chairs: Michela Becchi (North
  Carolina State University); John Feo (Pacific Northwest National Laborato
 ry (PNNL)); Antonino Tumeo (Pacific Northwest National Laboratory (PNNL));
  and Ana Lucia Varbanescu (University of Twente, Netherlands; University o
 f Amsterdam, Netherlands)\n\n
END:VEVENT
END:VCALENDAR