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:20240116T191700Z
LOCATION:DEF Concourse
DTSTART;TZID=America/Denver:20231114T100000
DTEND;TZID=America/Denver:20231114T170000
UID:submissions.supercomputing.org_SC23_sess291_rpost221@linklings.com
SUMMARY:Hybrid CPU-GPU Implementation of Edge-Connected Jaccard Similarity
  in Graph Datasets
DESCRIPTION:Posters, Research Posters\n\nAtharva Gondhalekar, Paul Sathre,
  and Wu-chun Feng (Virginia Tech)\n\nTypical GPU programs consist of four 
 steps: (1) data preparation, (2) host CPU-to-GPU data transfers, (3) execu
 tion of one or more GPU kernels, and (4) transfer of results back to CPU. 
 While the kernel is running on the GPU, the CPU cores often remain idle, w
 aiting on the GPU to finish  kernel execution.\n\nIn recent years, several
  frameworks have been presented that perform automated distribution of wor
 kload to both CPU and GPU. While the aforementioned frameworks offer techn
 iques for CPU+GPU workload distribution for regular applications, identify
 ing a performant CPU+GPU workload distribution for irregular applications 
 remains a difficult problem due to workload imbalance and irregular memory
  access patterns.\n\nThis work evaluates a hybrid CPU+GPU implementation o
 f an irregular workload -- graph link prediction using the Jaccard similar
 ity.  For the graphs that benefit the most from our hybrid CPU-GPU approac
 h, our implementation delivers a 16.4-28.4% improvement over the state-of-
 the-art Jaccard similarity implementation.\n\nRegistration Category: Tech 
 Program Reg Pass, Exhibits Reg Pass
END:VEVENT
END:VCALENDAR