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:20260422T143140Z
LOCATION:B306
DTSTART;TZID=America/New_York:20241118T105500
DTEND;TZID=America/New_York:20241118T112000
UID:submissions.supercomputing.org_SC24_sess751_ws_p3hpc114@linklings.com
SUMMARY:High-Performance, Scalable Geometric Multigrid via Fine-Grain Data
  Blocking for GPUs
DESCRIPTION:Oscar Antepara, Samuel Williams, and Hans Johansen (Lawrence B
 erkeley National Laboratory (LBNL)) and Mary Hall (University of Utah)\n\n
 We present a performance study of geometric multigrid (GMG) on NVIDIA, AMD
 , and Intel GPU-accelerated supercomputers.  The approach employs fine-gra
 in data blocking in BrickLib, which reduces data movement in the GMG V-cyc
 le by optimizing storage order for stencil access and communication.\nOur 
 GMG attains 73% in a peak performance portability metric, and 87% parallel
  efficiency when weak scaling to 512 GPUs on all three GPU-accelerated sup
 ercomputers.\nAnalysis shows stencil performance and MPI communication is 
 well-correlated with a traditional linear model from which we can extract 
 empirical latency, overhead, bandwidth, and throughput for comparison to t
 heoretical GPU and network limits.\nObservations show NVIDIA GPUs provide 
 the lowest overhead and highest throughput per process with AMD and Intel 
 GPUs delivering comparable performance.\nConversely, despite all three pla
 tforms employing the same Slingshot network, sustained bandwidth and laten
 cy vary widely when each GPU is dedicated one NIC.\n\nTag: Performance Opt
 imization, Programming Frameworks and System Software\n\nRegistration Cate
 gory: Workshop Reg Pass\n\nSession Chairs: CJ Newburn (NVIDIA Corporation)
 , Scott J. Parker (Argonne National Laboratory (ANL)), John Pennycook (Int
 el Corporation), and Kenneth Weiss (Lawrence Livermore National Laboratory
  (LLNL))\n\n
END:VEVENT
END:VCALENDAR