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:20260202T201247Z
LOCATION:Second Floor Atrium
DTSTART;TZID=America/Chicago:20251120T080000
DTEND;TZID=America/Chicago:20251120T170000
UID:submissions.supercomputing.org_SC25_sess533_post203@linklings.com
SUMMARY:Multi-GPU Implementation and Roofline Analysis of a Numerical Glob
 al Ocean Model
DESCRIPTION:Takateru Yamagishi (Research Organization for Information Scie
 nce and Technology); Masao Kurogi and Takao Kawasaki (Japan Agency for Mar
 ine-Earth Science and Technology); Yoshimasa Matsumura (National Institute
  for Environmental Studies); and Hiroyasu Hasumi (Atmosphere and Ocean Res
 earch Institute, The University of Tokyo)\n\nNumerical ocean models are es
 sential tools for climate prediction and marine resource studies, requirin
 g high resolution and realistic physical processes. We developed the globa
 l ocean model COCO and implemented it on GPUs using an OpenACC directive-b
 ased approach, while maintaining compatibility with CPUs. Performance was 
 evaluated on the Miyabi supercomputer, which includes GPU-based (NVIDIA GH
 200) and CPU-based (Intel Xeon MAX 9480) systems. Realistic ocean experime
 nts with a 0.17° global grid showed that most components achieved faster e
 xecution on GPUs, with the tracer calculation accelerated by a factor of 2
 .9. Roofline analysis revealed that most loops were memory-bound, and GPU 
 speedup was constrained by memory bandwidth rather than compute capability
 . Future improvements will require increasing arithmetic intensity and app
 lying kernel-level optimizations, while ensuring compatibility between CPU
 - and GPU-based codes.\n\nTag: Research & ACM SRC Posters\n\nRegistration 
 Category: Technical Program Reg Pass\n\nSession Chairs: Kento Sato (RIKEN 
 Center for Computational Science (R-CCS)); Chris Schlipalius (Pawsey Super
 computing Research Centre; Commonwealth Scientific and Industrial Research
  Organisation (CSIRO), Australia); and Anja Gerbes (Georg-August-Universit
 ät Göttingen)\n\n
END:VEVENT
END:VCALENDAR
