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:20250626T234542Z
LOCATION:B309
DTSTART;TZID=America/New_York:20241122T103000
DTEND;TZID=America/New_York:20241122T105000
UID:submissions.supercomputing.org_SC24_sess772_ws_hpct112@linklings.com
SUMMARY:Understanding Data Movement in AMD Multi-GPU Systems with Infinity
  Fabric
DESCRIPTION:Gabin Schieffer, Ruimin Shi, and Stefano Markidis (KTH Royal I
 nstitute of Technology); Andreas Herten (Forschungszentrum Jülich); and Je
 nnifer Faj and Ivy Peng (KTH Royal Institute of Technology)\n\nModern GPU 
 systems are constantly evolving to meet the needs of computing-intensive a
 pplications in scientific and machine learning domains. However, there is 
 typically a gap between hardware capacity and achievable application perfo
 rmance. This work aims to provide a better understanding of the Infinity F
 abric interconnects on AMD GPUs and CPUs. We propose a test and evaluation
  methodology for characterizing performance of data movements on multi-GPU
  systems, stressing different communication options on AMD MI250X GPUs, in
 cluding point-to-point and collective communication, and memory allocation
  strategies between GPUs, as well as the host CPU. In a single-node setup 
 with four GPUs, we show that direct peer-to-peer memory accesses between G
 PUs and utilization of the RCCL library outperform MPI-based solutions in 
 terms of memory/communication latency and bandwidth. Our test and evaluati
 on method serves as a base for validating memory and communication strateg
 ies on a system and improving applications on AMD multi-GPU computing syst
 ems.\n\nTag: Debugging and Correctness Tools, Hardware Technologies, Resou
 rce Management, State of the Practice\n\nRegistration Category: Workshop R
 eg Pass\n\nSession Chairs: Bilel Hadri (King Abdullah University of Scienc
 e and Technology (KAUST)) and Verónica G. Melesse Vergara (Oak Ridge Natio
 nal Laboratory (ORNL))\n\n
END:VEVENT
END:VCALENDAR