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:20210916T132453Z
LOCATION:Mère Royaume
DTSTART;TZID=Europe/Stockholm:20210708T170000
DTEND;TZID=Europe/Stockholm:20210708T173000
UID:submissions.pasc-conference.org_PASC21_sess200_msa370@linklings.com
SUMMARY:Assessing Posit in HPC Scientific Applications
DESCRIPTION:Minisymposium\n\nAssessing Posit in HPC Scientific Application
 s\n\nMarkidis, Chien, Podobas, Jansson, Peng\n\nFloating-point operations 
 can significantly impact the accuracy and performance of scientific applic
 ations on large-scale parallel systems. Recently, the Posit format has att
 racted attention as an alternative to the standard IEEE floating-point for
 mats because it could enable higher precision than IEEE formats using the 
 same number of bits. In this work, we first explore the feasibility of Pos
 it encoding in representative HPC applications by providing a 32-bit Posit
  NAS Parallel Benchmark (NPB) suite. Then, we evaluate the accuracy improv
 ement in different HPC kernels compared to the IEEE 754 format. Our result
 s indicate that using Posit encoding achieves optimized precision, ranging
  from 0.6 to 1.4 decimal digit, for all tested kernels and proxy applicati
 ons. Also, we quantified the overhead of the current software implementati
 on of Posit encoding as 4x-19x that of IEEE 754 hardware implementation. O
 ur study highlights the potential of hardware implementations of Posit to 
 benefit a broad range of HPC applications.\n\nDomain: CS and Math, Emergin
 g Applications, Physics, Engineering
END:VEVENT
END:VCALENDAR