BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//cfp.riscv-europe.org//eu-summit-2026//speaker//HGSFVS
BEGIN:VTIMEZONE
TZID:CET
BEGIN:STANDARD
DTSTART:20001029T040000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20000326T030000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-eu-summit-2026-A77QAQ@cfp.riscv-europe.org
DTSTART;TZID=CET:20260609T142000
DTEND;TZID=CET:20260609T143000
DESCRIPTION:Efficient deployment of Deep Learning (DL) models on RISC-V-bas
 ed multi-core platforms remains a significant challenge\, especially when 
 multiple models with heterogeneous structures and precision requirements m
 ust run concurrently. Existing frameworks offer optimized execution for si
 ngle-model inference but lack support for multi-model scheduling\, as well
  as priority-based resource allocation.\nIn this work\, we extend the capa
 bilities of such frameworks by formalizing the problem of multi-model\, mu
 ltiprecision inference scheduling on constrained many-core architectures l
 ike Parallel Ultra-Low Power (PULP). We define a scheduling space where mu
 ltiple Deep Neural Networks (DNNs)\, varying in size\, type and precision\
 , compete for limited computing and memory resources. We introduce a simpl
 e\, priority-aware scheduling layer that allocates cores and memory tiles 
 across models\, aiming to either minimize overall inference latency or fin
 d a tradeoff satisfying each model’s deadline.\nTo demonstrate the effec
 tiveness of our approach\, we leverage the existing Deployment Oriented to
  memoRY (DORY) framework\, and apply a greedy scheduling strategy. We cond
 ucted experiments with several models across several tasks and showed that
  even basic scheduling policies can significantly improve latency\, core u
 tilization\, and memory efficiency over static and sequential baselines.
DTSTAMP:20260522T162348Z
LOCATION:Poster Island C
SUMMARY:Priority-Aware Scheduling of Multi-Model\, Multi-Precision DNN Infe
 rence on Multi-Cores RISC-V - PGA\, GARREAU
URL:https://cfp.riscv-europe.org/eu-summit-2026/talk/A77QAQ/
END:VEVENT
END:VCALENDAR