Neural Heterogeneous Scheduler
This addresses scheduling efficiency for developers and researchers using next-generation heterogeneous SoCs, representing an incremental improvement over existing heuristics.
The paper tackles the problem of scheduling tasks on heterogeneous system-on-chips (SoCs) with multiple coprocessors, introducing a deep reinforcement learning method that learns to schedule tasks to minimize execution time, and empirically shows it outperforms conventional scheduling heuristics.
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.