BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20210916T132452Z
LOCATION:Mère Royaume
DTSTART;TZID=Europe/Stockholm:20210707T153000
DTEND;TZID=Europe/Stockholm:20210707T160000
UID:submissions.pasc-conference.org_PASC21_sess155_msa360@linklings.com
SUMMARY:Properties of GMRES with Iterative Refinement on GPUs
DESCRIPTION:Minisymposium\n\nProperties of GMRES with Iterative Refinement
  on GPUs\n\nLoe, Glusa, Yamazaki, Boman, Rajamanickam\n\nAlgorithms to sol
 ve large, sparse linear systems Ax=b are crucial to many applications.  Th
 e popular GMRES algorithm is typically memory bound on modern parallel com
 puters.  Multiprecision algorithms reduce the cost of data movement by sto
 ring some data in a lower precision format.  We implement GMRES with itera
 tive refinement (GMRES-IR) and show performance improvements on GPUs over 
 double-precision GMRES. We combine GMRES-IR with several scalable precondi
 tioners, such as block Jacobi and polynomial preconditioning, to show addi
 tional improvements to solve time and performance.\n\nDomain: CS and Math
END:VEVENT
END:VCALENDAR