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:20260522T150120Z
LOCATION:D172
DTSTART;TZID=America/Chicago:20181112T162000
DTEND;TZID=America/Chicago:20181112T164500
UID:submissions.supercomputing.org_SC18_sess168_ws_ia107@linklings.com
SUMMARY:High-Performance GPU Implementation of PageRank with Reduced Preci
 sion Based on Mantissa Segmentation
DESCRIPTION:Thomas Grützmacher and Hartwig Anzt (Karlsruhe Institute of T
 echnology), Florian Scheidegger (ETH Zurich), and Enrique S. Quintana-Ortí
  (Jaume I University)\n\nWe address the acceleration of the PageRank algor
 ithm for web information retrieval on graphics processing units (GPUs) via
  a modular precision framework that adapts the input data format in memory
  to the numerical requirements as the iteration converges. In detail, we a
 bandon the ieee 754 single- and double-precision number representation for
 mats, employed in the standard implementation of PageRank, to instead stor
 e the data in memory in some specialized formats. Furthermore, we avoid th
 e data duplication by leveraging a data layout based on mantissa segmentat
 ion. Our evaluation on a V100 graphics card from NVIDIA shows acceleration
  factors of up to 30% with respect to the standard algorithm operating in 
 double-precision.\n\nTag: Architectures, Data Analytics, Graph Algorithms\
 n\nRegistration Category: Workshop Reg Pass\n\nSession Chairs: Vito Giovan
 ni Castellana (Pacific Northwest National Laboratory (PNNL)), John Feo (Pa
 cific Northwest National Laboratory (PNNL)), and Antonino Tumeo (Pacific N
 orthwest National Laboratory (PNNL))\n\n
END:VEVENT
END:VCALENDAR