Overview

Real-time scheduling decides when and where tasks run so control loops, data flows, and user requests finish within their deadlines. Common approaches include fixed-priority and earliest-deadline-first policies, DAG and workflow schedulers, resource-aware dispatching, and optimization or reinforcement learning; profiling and schedulability analysis help estimate whether deadlines can be met. Applications range from embedded and robotic systems to cloud workflows, industrial production, and machine-learning inference, where deadlines may compete with throughput, energy, accuracy, and fairness goals. A central challenge is maintaining predictable timing under changing workloads and shared-resource interference without making online decisions too costly or sacrificing efficiency.

Latest papers

CardsList

34 papers5 papers in the last four weeks, up 25% on the four weeks before. 0.0% of all new papers.

  1. FlowPrefill: Decoupling Preemption from Prefill Scheduling Granularity to Mitigate Head-of-Line Blocking in LLM Serving

    Feb 18, 2026Chia-chi Hsieh, Zan Zong, Xinyang Chen +3LLM ServingLLM Inference Scheduling

  2. HiDVFS: Hierarchical Multi-Agent DVFS for Real-Time OpenMP DAG Workloads

    Jan 10, 2026Mohammad Pivezhandi, Abusayeed Saifullah, Ali JannesariReal-Time SchedulingMulti-Agent Reinforcement Learning

  3. AEGIS: Runtime-Guided GPU Collocation for Multi-Tenant Deep Learning Training

    Aug 26, 2025Ehsan Yousefzadeh-Asl-Miandoab, Büşra Karatay Demiray, Florina M. Ciorba +2Real-Time SchedulingOnline Resource Allocation