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:20230124T171527Z
LOCATION:C144-145
DTSTART;TZID=America/Chicago:20221114T154500
DTEND;TZID=America/Chicago:20221114T160000
UID:submissions.supercomputing.org_SC22_sess462_ws_ai4s101@linklings.com
SUMMARY:Practical Federated Learning Infrastructure for Privacy-Preserving
  Scientific Computing
DESCRIPTION:Workshop\n\nPractical Federated Learning Infrastructure for Pr
 ivacy-Preserving Scientific Computing\n\nWang, Zhao\n\nFederated learning 
 (FL) is deemed a promising paradigm for privacy-preserving data analytics 
 in collaborative scientific computing.\nHowever, there lacks an effective 
 and easy-to-use FL infrastructure for scientific computing and high-perfor
 mance computing (HPC) environments.  The objective of this presentation is
  two-fold.  First, we identify three missing pieces of a scientific FL inf
 rastructure: \n   (i) a native MPI programming interface that can be seaml
 essly integrated into existing scientific applications, \n   (ii) an indep
 endent data layer for the FL system such that the user can pick the persis
 tent medium for her own choice, such as parallel file systems and multidim
 ensional databases, \n   (iii) efficient encryption protocols that are opt
 imized for scientific workflows. \n\nThe second objective is to present a 
 work-in-progress FL infrastructure, namely MPI-FL, which is implemented wi
 th PyTorch and MPI4py.  We deploy MPI-FL on 1,000 CPU cores and evaluate i
 t with four standard benchmarks: MNIST, Fashion-MNIST, CIFAR-10, and SVHN-
 extra.\n\nSession Format: Recorded\n\nRegistration Category: Workshop Reg 
 Pass
END:VEVENT
END:VCALENDAR