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:20251116T163000
DTEND;TZID=America/Chicago:20251116T165000
UID:submissions.supercomputing.org_SC25_sess201_ws_memo104@linklings.com
SUMMARY:Hardware-Software Co-Design of Iterative Filter-Update Numerical M
 ethods Using Processing-In-Memory
DESCRIPTION:Eric Tang and Tianyun Zhang (Carnegie Mellon University), Will
 iam Bradford and Farzana Ahmed Siddique (University of Virginia), James C.
  Hoe (Carnegie Mellon University), Kevin Skadron (University of Virginia),
  and Franz Franchetti (Carnegie Mellon University)\n\nData movement is a k
 ey bottleneck in applications such as machine learning and scientific comp
 uting. Some software techniques address this by computing on subsets of da
 ta but this still requires reading the entire dataset to determine the sub
 set. We propose a hardware-software co-design approach for iterative metho
 ds centered around two operations--filtering and updating. We introduce a 
 domain-specific language that supports these computational patterns to ena
 ble PIM programming. Since filter and update are simple pointwise operatio
 ns, PIM hardware requires only limited compute capability.\n\nIn this work
 , we investigate gradient descent for an ill-conditioned convex optimizati
 on function using this approach and map it to a PIM architecture using the
  PIMEval architectural simulator. Filter load and update store operations 
 sparsify the data set by 83% while requiring as few as 1.5x more iteration
 s to converge compared to traditional gradient descent approaches, with a 
 net reduction in data movement of 3.9x.\n\nRecording: Livestreamed, Record
 ed\n\nRegistration Category: Technical Program Reg Pass, Workshop Reg Pass
 \n\nSession Chairs: Stephen L. Olivier (Sandia National Laboratories), May
 a Gokhale (Lawrence Livermore National Laboratory (LLNL)), Ivy Peng (KTH R
 oyal Institute of Technology), Kyle Hale (Oregon State University), and Ro
 nald Minnich (Hewlett Packard Enterprise (HPE))\n\n
END:VEVENT
END:VCALENDAR