BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Chicago
X-LIC-LOCATION:America/Chicago
BEGIN:DAYLIGHT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
TZNAME:CDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
TZNAME:CST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260522T150116Z
LOCATION:A2 Ballroom
DTSTART;TZID=America/Chicago:20181115T110000
DTEND;TZID=America/Chicago:20181115T113000
UID:submissions.supercomputing.org_SC18_sess467_gb106@linklings.com
SUMMARY:ShenTu: Processing Multi-Trillion Edge Graphs on Millions of Cores
  in Seconds
DESCRIPTION:Heng Lin (Tsinghua University, Fma Technology); Xiaowei Zhu (T
 singhua University, Qatar Computing Research Institute); Bowen Yu (Tsinghu
 a University); Xiongchao Tang (Tsinghua University, Qatar Computing Resear
 ch Institute); Wei Xue and Wenguang Chen (Tsinghua University); Lufei Zhan
 g (State Key Laboratory of Mathematical Engineering and Advanced Computing
 ); Torsten Hoefler (ETH Zurich); Xiaosong Ma (Qatar Computing Research Ins
 titute); Xin Liu (National Research Centre of Parallel Computer Engineerin
 g and Technology); Weimin Zheng (Tsinghua University); and Jingfang Xu (Be
 ijing Sogou Technology Development Company)\n\nGraphs are an important abs
 traction used in many scientific fields. With the magnitude of graph-struc
 tured data constantly increasing, effective data analytics requires effici
 ent and scalable graph processing systems. Although HPC systems have long 
 been used for scientific computing, people have only recently started to a
 ssess their potential for graph processing, a workload with inherent load 
 imbalance, lack of locality, and access irregularity. We propose ShenTu, t
 he first general-purpose graph processing framework that can efficiently u
 tilize an entire petascale system to process multi-trillion edge graphs in
  seconds. ShenTu embodies four key innovations: hardware specializing, sup
 ernode routing, on-chip sorting, and degree-aware messaging, which togethe
 r enable its unprecedented performance and scalability. It can traverse an
  unprecedented 70-trillion-edge graph in seconds. Furthermore, ShenTu enab
 les the processing of a spam detection problem on a 12-trillion edge Inter
 net graph, making it possible to identify trustworthy and spam web pages d
 irectly at the fine-grained page level.\n\n
END:VEVENT
END:VCALENDAR