BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/New_York
X-LIC-LOCATION:America/New_York
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260422T143139Z
LOCATION:B206
DTSTART;TZID=America/New_York:20241119T133000
DTEND;TZID=America/New_York:20241119T140000
UID:submissions.supercomputing.org_SC24_sess679_exforum129@linklings.com
SUMMARY:Accelerating Scientific Computing: GPU Optimization Strategies, Ch
 allenges, and Performance Outcomes
DESCRIPTION:Rongguang Jia and Alex Main (ANSYS)\n\nGPUs are transforming s
 cientific computing by delivering substantial speedups and energy savings.
  This presentation outlines the development of GPU computing strategies fo
 r a leading CFD software. Initially, we identified and optimized computati
 onal bottlenecks suitable for GPU acceleration, using an offload model. To
  overcome limitations imposed by Amdahl's law, we developed a GPU-native s
 olver architecture with streamlined APIs, ensuring seamless integration wi
 th existing workflows. \n\nOur optimization strategy also accounts for div
 erse GPU platforms, implementing platform-specific enhancements for NVIDIA
 , AMD, and Intel architectures. Scalability was achieved through advanced 
 load-balancing algorithms and improved inter-GPU communication, enabling e
 fficient parallelization for large-scale simulations. \n\nWe present resul
 ts demonstrating significant speedups and energy savings compared to CPU-b
 ased methods, highlighting the transformative potential of GPUs in enablin
 g faster, more complex simulations.\n\nTag: Accelerators, Software Enginee
 ring\n\nRegistration Category: Tech Program Reg Pass, Exhibits Reg Pass\n\
 nSession Chair: Scott Michael (Indiana University)\n\n
END:VEVENT
END:VCALENDAR