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:20210402T160554Z
LOCATION:Track 10
DTSTART;TZID=America/New_York:20201111T113000
DTEND;TZID=America/New_York:20201111T115000
UID:submissions.supercomputing.org_SC20_sess203_ws_scc103@linklings.com
SUMMARY:Performance Characteristics of Virtualized GPUs for Deep Learning
DESCRIPTION:Workshop\n\nPerformance Characteristics of Virtualized GPUs fo
 r Deep Learning\n\nMichael, Teige, Li, Lowe, Turner...\n\nAs deep learning
  techniques and algorithms become more and more common in scientific workf
 lows, HPC centers are grappling with how best to provide GPU resources and
  support deep learning workloads.  One novel method of deployment is to vi
 rtualize GPU resources allowing for multiple VM instances to have logicall
 y distinct virtual GPUs (vGPUs) on a shared physical GPU.  There are many 
 operational and performance implications to consider, however, before depl
 oying a vGPU service in an HPC center.  In this paper, we investigate the 
 performance characteristics of vGPUs for both traditional HPC workloads an
 d for deep learning training and inference workloads. Using NVIDIA's vDWS 
 virtualization software, we perform a series of HPC and deep learning benc
 hmarks on both non-virtualized (bare metal) and vGPUs of various sizes and
  configurations.  We report on several of the challenges we discovered in 
 deploying and operating a variety of virtualized instance sizes and config
 urations.  We find that the overhead of virtualization on HPC workloads is
  generally less than 10%, and can vary considerably for deep learning, dep
 ending on the task.\n\nTag: Big Data, Cloud and Distributed Computing, Ext
 reme Scale Computing, Scientific Computing, Workflows\n\nRegistration Cate
 gory: Workshop Reg Pass
END:VEVENT
END:VCALENDAR