Neural Heterogeneous Scheduler

2019/06/09 by Tegg Taekyong Sung, Tegg Tae Kyong Sung, Valliappa Chockalingam +6
Computer Science · Engineering · #Advanced Memory and Neural Computing #FOS: Computer and information sciences #Machine Learning (cs.LG) #Operating Systems (cs.OS) #Parallel Computing and Optimization Techniques #Reinforcement Learning in Robotics #cs.LG #cs.OS

paper · pdf · doi:10.48550/arxiv.1906.03724

7 pages. The first two authors contributed equally. ICML 2019 Real-world Sequential Decision Making Workshop

arxiv created 2019/06/09 · openalex publication_date 2019/06/09 · arxiv updated 2019/06/11 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Access to parallel and distributed computation has enabled researchers and developers to improve algorithms and performance in many applications. Recent research has focused on next generation special purpose systems with multiple kinds of coprocessors, known as heterogeneous system-on-chips (SoC). In this paper, we introduce a method to intelligently schedule--and learn to schedule--a stream of tasks to available processing elements in such a system. We use deep reinforcement learning enabling complex sequential decision making and empirically show that our reinforcement learning system provides for a viable, better alternative to conventional scheduling heuristics with respect to minimizing execution time.

Related

• Deep Reinforcement Learning for System-on-Chip: Myths and Realities
• SoCRATES: System-on-Chip Resource Adaptive Scheduling using Deep Reinforcement Learning
• Real-Time Scheduling of Machine Learning Operations on Heterogeneous Neuromorphic SoC
• HeterPS: Distributed Deep Learning With Reinforcement Learning Based Scheduling in Heterogeneous Environments
• Geometric Deep Reinforcement Learning for Dynamic DAG Scheduling