SoCRATES: System-on-Chip Resource Adaptive Scheduling using Deep Reinforcement Learning

TL;DR

SoCRATES is a deep reinforcement learning-based scheduler for System-on-Chip that minimizes average latency for hierarchical, heterogeneous tasks, outperforming existing methods with significant efficiency and energy benefits.

Contribution

Introduces SoCRATES, a novel DRL-based scheduler that effectively minimizes latency in SoC resource management, addressing task dependency challenges with high efficiency.

Findings

Achieves up to 38% latency reduction compared to existing schedulers.

Operates efficiently with a compact model size and low energy consumption.

Effectively handles task dependencies in latency-sensitive applications.

Abstract

Deep Reinforcement Learning (DRL) is being increasingly applied to the problem of resource allocation for emerging System-on-Chip (SoC) applications, and has shown remarkable promises. In this paper, we introduce SoCRATES (SoC Resource AdapTivE Scheduler), an extremely efficient DRL-based SoC scheduler which maps a wide range of hierarchical jobs to heterogeneous resources within SoC using the Eclectic Interaction Matching (EIM) technique. It is noted that the majority of SoC resource management approaches have been targeting makespan minimization with fixed number of jobs in the system. In contrast, SoCRATES aims at minimizing average latency in a steady-state condition while assigning tasks in the ready queue to heterogeneous resources (processing elements). We first show that the latency-minimization-driven SoC applications operate high-frequency job workload and distributed/parallel job execution. We then demonstrate SoCRATES successfully addresses the challenge of concurrent observations caused by the task dependency inherent in the latency minimization objective. Extensive tests show that SoCRATES outperforms other existing neural and non-neural schedulers with as high as 38% gain in latency reduction under a variety of job types and incoming rates. The resulting model is also compact in size and has very favorable energy consumption behaviors, making it highly practical for deployment in future SoC systems with built-in neural accelerator.