BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Los_Angeles
X-LIC-LOCATION:America/Los_Angeles
BEGIN:DAYLIGHT
TZOFFSETFROM:-0800
TZOFFSETTO:-0700
TZNAME:PDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0700
TZOFFSETTO:-0800
TZNAME:PST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260730T152640Z
LOCATION:Exhibit Hall
DTSTART;TZID=America/Los_Angeles:20260728T174900
DTEND;TZID=America/Los_Angeles:20260728T175000
UID:dac_DAC 2026_sess306_WIP3277@linklings.com
SUMMARY:RELMAS-HRT: Online Inference Scheduling in Mixed-Criticality Multi
 -Tenant Multi-Accelerator Systems via Reinforcement Learning
DESCRIPTION:Francesco Blanco and Enrico Russo (Univerisity of Catania) and
  Giuseppe Ascia and Maurizio Palesi (University of Catania)\n\nWe present 
 RELMAS-HRT, an online reinforcement learning scheduler for heterogeneous m
 ulti-accelerator systems (MAS) that guarantees hard real-time (HRT) deadli
 nes while optimizing QoS for multi-tenant DNN inference. The framework adm
 its HRT tasks through a WCET-based feasibility test, constructs a determin
 istic baseline schedule, and dynamically refines it at runtime using an RL
  policy whose decisions are accepted only when provably safe. RELMAS-HRT e
 xploits slack and missing HRT releases to improve QoS-aware throughput wit
 hout compromising guarantees. Experiments show 100% HRT deadline satisfact
 ion and higher QoS adherence with respect to baselines policies, enabling 
 efficient mixed-criticality inference in industrial systems.\n\nTrack: Stu
 dent\n\n
END:VEVENT
END:VCALENDAR