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:C141/143/149
DTSTART;TZID=America/Chicago:20181114T163000
DTEND;TZID=America/Chicago:20181114T170000
UID:submissions.supercomputing.org_SC18_sess215_pap503@linklings.com
SUMMARY:ADAPT: Algorithmic Differentiation Applied to Floating-Point Preci
 sion Tuning
DESCRIPTION:Harshitha Menon (Lawrence Livermore National Laboratory); Mich
 ael O. Lam (James Madison University, Lawrence Livermore National Laborato
 ry); and Daniel Osei-Kuffuor, Markus Schordan, Scott Lloyd, Kathryn Mohror
 , and Jeffrey Hittinger (Lawrence Livermore National Laboratory)\n\nHPC ap
 plications extensively use floating point arithmetic operations to solve c
 omputational problems in various domains. Mixed precision computing, use o
 f lowest precision data type sufficient to achieve a desired accuracy, hav
 e been explored to improve performance, reduce power consumption and data 
 movement. Manually optimizing the program to use mixed precision is challe
 nging. In this work, we present ADAPT, an approach for mixed precision ana
 lysis on HPC workloads while providing guarantees about the final output e
 rror. Our approach uses algorithmic differentiation to accurately estimate
  the output error for mixed precision configuration. ADAPT provides floati
 ng-point precision sensitivity of programs, which highlights regions of th
 e code that that can potentially be converted to lower precision, is used 
 to make algorithmic choices and develop mixed precision configurations. We
  evaluate ADAPT on six benchmarks and a proxy application and show that we
  are able to achieve a speedup of 1.2x on the proxy application, LULESH.\n
 \nTag: Algorithms, Applications, Architectures, Compiler Analysis and Opti
 mization, Floating Point, Performance, Precision, Programming Systems, Too
 ls\n\nRegistration Category: Tech Program Reg Pass\n\nSession Chair: Proto
 nu Basu (NVIDIA)\n\n
END:VEVENT
END:VCALENDAR