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:20181221T160728Z
LOCATION:D165
DTSTART;TZID=America/Chicago:20181112T160000
DTEND;TZID=America/Chicago:20181112T163000
UID:submissions.supercomputing.org_SC18_sess161_ws_pmbsf114@linklings.com
SUMMARY:Approximating a Multi-Grid Solver
DESCRIPTION:Workshop\nBenchmarks, Parallel Programming Languages, Librarie
 s, and Models, Performance, Simulation, Workshop Reg Pass\n\nApproximating
  a Multi-Grid Solver\n\nLe Fèvre, Bautista-Gomez, Unsal, Casas\n\nMulti-gr
 id methods are numerical algorithms used in parallel and distributed proce
 ssing. The main idea of multi-grid solvers is to speed up the convergence 
 of an iterative method by reducing the problem to a coarser grid a number 
 of times. Multi-grid methods are widely exploited in many application doma
 ins, thus it is important to improve their performance and energy efficien
 cy. This paper aims to reach this objective based on the following observa
 tion: Given that the intermediary steps do not require full accuracy, it i
 s possible to save time and energy by reducing precision during some steps
  while keeping the final result within the targeted accuracy.\n\nTo achiev
 e this goal, we first introduce a cycle shape different from the classic V
 -cycle used in multi-grid solvers.  Then, we propose to dynamically change
  the floating-point precision used during runtime according to the accurac
 y needed for each intermediary step. Our evaluation considering a state-of
 -the-art multi-grid solver implementation demonstrates that it is possible
  to trade temporary precision for time to completion without hurting the q
 uality if the final result.  In particular, we are able to reach the same 
 accuracy results as with full double-precision while gaining between 15% a
 nd 30% execution time improvement.
URL:https://sc18.supercomputing.org/presentation/?id=ws_pmbsf114&sess=sess
 161
END:VEVENT
END:VCALENDAR