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:20260202T201258Z
LOCATION:Second Floor Atrium
DTSTART;TZID=America/Chicago:20251121T080000
DTEND;TZID=America/Chicago:20251121T120000
UID:submissions.supercomputing.org_SC25_sess620_post297@linklings.com
SUMMARY:Local vs. Global FFT Approaches for High-Performance Ultrasound Si
 mulation on Multi-GPU Systems
DESCRIPTION:Oliver Kuník and Jiri Jaros (Faculty of Information Technology
 , Brno University of Technology)\n\nSimulating wave propagation with the F
 ourier collocation method is computationally intensive due to its reliance
  on discrete Fourier transforms (DFTs). While DFTs enable near-minimal spa
 tial discretization, they scale poorly on modern high performance computin
 g systems. This work evaluates two multi-GPU strategies for three-dimensio
 nal simulations: a Global FFT approach using distributed transforms, and a
  Local FFT approach based on domain decomposition with halo exchanges. Exp
 eriments were performed on a system with eight NVIDIA A100 GPUs connected 
 via NVSwitch. Precision tests show that the Local FFT approach maintains e
 rrors around 0.1% when the halo covers the local PML region. Performance r
 esults demonstrate that the Local FFT approach achieves lower runtimes and
  significantly reduced communication overhead compared to the Global FFT a
 pproach, particularly for larger domains. These findings indicate that Loc
 al FFT decomposition is a promising strategy for scalable, large-scale mul
 ti-node ultrasound simulations.\n\nTag: Research & ACM SRC Posters\n\nRegi
 stration Category: Technical Program Reg Pass\n\nSession Chairs: Kento Sat
 o (RIKEN Center for Computational Science (R-CCS)); Anja Gerbes (Georg-Aug
 ust-Universität Göttingen); and Chris Schlipalius (Pawsey Supercomputing R
 esearch Centre; Commonwealth Scientific and Industrial Research Organisati
 on (CSIRO), Australia)\n\n
END:VEVENT
END:VCALENDAR