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:20260202T201809Z
LOCATION:274
DTSTART;TZID=America/Chicago:20251117T113000
DTEND;TZID=America/Chicago:20251117T115000
UID:submissions.supercomputing.org_SC25_sess215_ws_ai4s118@linklings.com
SUMMARY:InferCT: An Efficient and Generalizable Framework to Enable 3D Mac
 hine Learning for Computed Tomography
DESCRIPTION:Austin Yunker, Weijian Zheng, and Rajkumar Kettimuthu (Argonne
  National Laboratory (ANL))\n\nIn this paper, we propose inferCT, an effic
 ient framework that enables 3D deep learning for computed tomography (CT) 
 during inference. Our baseline approach addresses this issue by partitioni
 ng CT volumes into cubic sub-volumes that fit into GPU memory and distribu
 ting them across multiple GPUs. Building on this, we introduce further ven
 dor-agnostic optimizations, including a lock-free shared memory data struc
 ture to reduce synchronization overhead, pipeline execution to hide data p
 refetching and post-processing latency, and a parallel data loader to impr
 ove I/O efficiency. Results on both AMD and NVIDIA GPUs show that our opti
 mized framework achieves speedups of 1.97× and 2.32× over the baseline for
  the 10243 and 40963 datasets, respectively. For the scalability tests, ex
 periments demonstrate strong scaling efficiencies of 89.25% and 75.75% whe
 n scaling from 1 to 4 GPUs within a single NUMA node, and from 1 to 8 GPUs
  across two NUMA nodes, respectively, using the 40963 dataset.\n\nRecordin
 g: Livestreamed, Recorded\n\nRegistration Category: Technical Program Reg 
 Pass, Workshop Reg Pass\n\nSession Chairs: Gokcen Kestor (Barcelona Superc
 omputing Center (BSC); University of California, Merced); Dong Li (Univers
 ity of California, Merced); and Murali Emani (Argonne National Laboratory 
 (ANL))\n\n
END:VEVENT
END:VCALENDAR