BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Denver
X-LIC-LOCATION:America/Denver
BEGIN:DAYLIGHT
TZOFFSETFROM:-0700
TZOFFSETTO:-0600
TZNAME:MDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0600
TZOFFSETTO:-0700
TZNAME:MST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260422T000711Z
LOCATION:301-302-303
DTSTART;TZID=America/Denver:20231114T153000
DTEND;TZID=America/Denver:20231114T160000
UID:submissions.supercomputing.org_SC23_sess167_pap185@linklings.com
SUMMARY:BLAD: Adaptive Load Balanced Scheduling and Operator Overlap Pipel
 ine for Accelerating the Dynamic GNN Training
DESCRIPTION:Kaihua Fu, Quan Chen, Yuzhuo Yang, Jiuchen Shi, Chao Li, and M
 inyi Guo (Shanghai Jiao Tong University)\n\nDynamic graph networks are wid
 ely used for learning time-evolving graphs, but prior work on training the
 se networks is inefficient due to communication overhead, long synchroniza
 tion, and poor resource usage.  Our investigation shows that communication
  and synchronization can be reduced by carefully scheduling the workload a
 nd the execution order of operators in GNNs can be adjusted without hurtin
 g training convergence.  \n\nWe propose a system called BLAD to consider t
 he above factors, comprising a two-level load scheduler and an overlap-awa
 re topology manager.  The scheduler allocates each snapshot group to a GPU
 , alleviating cross-GPU communication. \nThe snapshots in a group are then
  carefully allocated to processes on a GPU, enabling overlap of compute-in
 tensive NN operators and memory-intensive graph operators.  The topology m
 anager adjusts the operators' execution order to maximize the overlap.  Ex
 periments show that it achieves 27.2% speed up on training time on average
  without affecting final accuracy, compared to state-of-the-art solutions.
 \n\nTag: Artificial Intelligence/Machine Learning\n\nRegistration Category
 : Tech Program Reg Pass\n\nReproducibility Badges: Artifact Available\n\nS
 ession Chair: Israt Nisa (Amazon Web Services AI Research and Education)\n
 \n
END:VEVENT
END:VCALENDAR