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:20260202T201805Z
LOCATION:274
DTSTART;TZID=America/Chicago:20251116T161000
DTEND;TZID=America/Chicago:20251116T163000
UID:submissions.supercomputing.org_SC25_sess201_ws_memo103@linklings.com
SUMMARY:CaRDS: Compiler-aided Remote Data Structures
DESCRIPTION:Brian Tauro and Ian Dougherty (Illinois Tech) and Kyle Hale (O
 regon State University)\n\nFar memory tiers improve memory utilization by 
 enabling memory intensive applications to use idle memory from other machi
 nes over the network. Recently, compiler approaches to far memory have dem
 onstrated how static analysis can be leveraged to automatically transform 
 applications to make efficient use of remote memory tiers. However, polici
 es in these compilers, e.g., the determination of whether objects should b
 e remoted, prefetched, or evacuated are made conservatively at compile tim
 e or require profiling. While profiling can alleviate conservative policie
 s, profile-guided systems can be expensive and may not work well for appli
 cations that have variation in their inputs. We propose CaRDS, system that
  combines both runtime and static analysis to determine far memory policie
 s dynamically, at data structure granularity, and without profiling. CaRDS
  remoting policies can outperform prior automatic approaches by up to 2× a
 nd are within 25% of profile-guided systems when the local memory is highl
 y constrained.\n\nRecording: Livestreamed, Recorded\n\nRegistration Catego
 ry: Technical Program Reg Pass, Workshop Reg Pass\n\nSession Chairs: Steph
 en L. Olivier (Sandia National Laboratories), Maya Gokhale (Lawrence Liver
 more National Laboratory (LLNL)), Ivy Peng (KTH Royal Institute of Technol
 ogy), Kyle Hale (Oregon State University), and Ronald Minnich (Hewlett Pac
 kard Enterprise (HPE))\n\n
END:VEVENT
END:VCALENDAR