BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Denver
X-LIC-LOCATION:America/Denver
BEGIN:DAYLIGHT
TZOFFSETFROM:-0700
TZOFFSETTO:-0600
TZNAME:MDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0600
TZOFFSETTO:-0700
TZNAME:MST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20240116T191700Z
LOCATION:401-402
DTSTART;TZID=America/Denver:20231114T160000
DTEND;TZID=America/Denver:20231114T163000
UID:submissions.supercomputing.org_SC23_sess179_pap501@linklings.com
SUMMARY:HPAC-Offload: Accelerating HPC Applications with Portable Approxim
 ate Computing on the GPU
DESCRIPTION:Paper\n\nZane Fink (Lawrence Livermore National Laboratory (LL
 NL), University of Illinois) and Konstantinos Parasyris, Giorgis Georgakou
 dis, and Harshitha Menon (Lawrence Livermore National Laboratory (LLNL))\n
 \nThe end of Dennard scaling and the slowdown of Moore's law led to a shif
 t in technology trends toward parallel architectures, particularly in HPC 
 systems. To continue providing performance benefits, HPC should embrace Ap
 proximate Computing (AC), which trades application quality loss for improv
 ed performance. However, existing AC techniques have not been extensively 
 applied and evaluated in state-of-the-art hardware architectures such as G
 PUs, the primary execution vehicle for HPC applications today.\n\nThis pap
 er presents HPAC-Offload, a pragma-based programming model that extends Op
 enMP offload applications to support AC techniques, allowing portable appr
 oximations across different GPU architectures. We conduct a comprehensive 
 performance analysis of HPAC-Offload across GPU-accelerated HPC applicatio
 ns, revealing that AC techniques can significantly accelerate HPC applicat
 ions (1.64x LULESH on AMD, 1.57x NVIDIA) with minimal quality loss (0.1%).
  Our analysis offers deep insights into the performance of GPU-based AC th
 at guide the future development of AC algorithms and systems for these arc
 hitectures.\n\nTag: Accelerators, Distributed Computing, Middleware and Sy
 stem Software, Performance Measurement, Modeling, and Tools, Post-Moore Co
 mputing\n\nRegistration Category: Tech Program Reg Pass\n\nReproducibility
  Badges: Artifact Available\n\nSession Chair: Hari Subramoni (Ohio State U
 niversity)
END:VEVENT
END:VCALENDAR