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:20260422T143045Z
LOCATION:B312-B313A
DTSTART;TZID=America/New_York:20241119T153000
DTEND;TZID=America/New_York:20241119T170000
UID:submissions.supercomputing.org_SC24_sess375@linklings.com
SUMMARY:Leadership-class Supercomputers
DESCRIPTION:Exploring GPU-to-GPU Communication: Insights into Supercompute
 r Interconnects\n\nMulti-GPU nodes are increasingly common in the rapidly 
 evolving landscape of exascale supercomputers. On these systems, GPUs on t
 he same node are connected through dedicated networks, with bandwidths up 
 to a few terabits per second. However, gauging performance expectations an
 d maximizing system eff...\n\n\nDaniele De Sensi (Sapienza University of R
 ome); Lorenzo Pichetti and Flavio Vella (University of Trento, Italy); Tiz
 iano De Matteis and Zebin Ren (Vrije Universiteit Amsterdam); Luigi Fusco 
 (ETH Zürich); Matteo Turisini and Daniele Cesarini (CINECA); Kurt Lust (Un
 iversity of Antwerp, Belgium); Animesh Trivedi (IBM); Duncan Roweth (Hewle
 tt Packard Enterprise (HPE)); Filippo Spiga and Salvatore Di Girolamo (NVI
 DIA Corporation); and Torsten Hoefler (ETH Zürich)\n---------------------\
 nApplication-Driven Exascale: The JUPITER Benchmark Suite\n\nBenchmarks ar
 e essential in the design of modern HPC installations, as they define key 
 aspects of system components. Beyond synthetic workloads, it is crucial to
  include real applications that represent user requirements into benchmark
  suites, to guarantee high-usability and widespread adoption of a...\n\n\n
 Andreas Herten, Sebastian Achilles, Damian Alvarez, Jayesh Badwaik, Eric B
 ehle, Mathis Bode, Thomas Breuer, Daniel Caviedes Voullieme, Mehdi Cherti,
  Adel Dabah, Salem El Sayed, Wolfgang Frings, Ana Gonzales-Nicolas, Eric B
 . Gregory, Kaveh Haghighi Mood, Thorsten Hater, Jenia Jitsev, Chelsea Mari
 a John, Jan M. Meinke, Catrin Meyer, Pavel Mezentsev, Jan-Oliver Mirus, St
 epan Nassyr, Carolin Penke, Manoel Römmer, Ujjwal Sinha, Benedikt von St. 
 Vieth, and Olaf Stein (Forschungszentrum Jülich, Jülich Supercomputing Cen
 tre (JSC)); Estela Suarez (Forschungszentrum Jülich; University of Bonn, G
 ermany); and Dennis Willsch and Ilya Zhukov (Forschungszentrum Jülich, Jül
 ich Supercomputing Centre (JSC))\n---------------------\nAn Evaluation of 
 the Effect of Network Cost Optimization for Leadership Class Supercomputer
 s\n\nThe Dragonfly is an extensively deployed network\ntopology in large-s
 cale high-performance computing (HPC) due\nto its cost-effectiveness and e
 fficiency. In comparison to other in-\ndirect networks of similar scale, t
 he Dragonfly network has shown\na considerable reduction in cable lengths 
 and network c...\n\n\nAwais Khan, John Lange, Nick Hagerty, Edwin F. Posad
 a, John Holmen, James White, Austin Harris, Veronica Melesse Vergara, Chri
 stopher Zimmer, and Scott Atchley (Oak Ridge National Laboratory (ORNL))\n
 \nTag: Accelerators, HPC Infrastructure, Performance Evaluation and/or Opt
 imization Tools, State of the Practice\n\nRegistration Category: Tech Prog
 ram Reg Pass\n\nAward Finalist: Best Paper Finalist\n\nSession Chair: Zach
 ary Tschirhart (Sustainment)
END:VEVENT
END:VCALENDAR