A Survey of Real-time Scheduling on Accelerator-based Heterogeneous Architecture for Time Critical Applications

2025-05-17

AI Summary

This paper addresses the stringent timing guarantees required for heterogeneous CPU–accelerator (GPU/TPU/FPGA) architectures in real-time robotics and autonomous driving systems. It presents a systematic survey of soft- and hard-real-time scheduling research from 2014 to 2024. The authors propose a unified modeling framework that jointly captures heterogeneous hardware characteristics (e.g., memory hierarchies, interconnects) and task execution behaviors (e.g., kernel launch latency, stream dependencies). They introduce the first holistic taxonomy covering vendor-specific runtime abstractions (CUDA Streams, Vitis), response-time analysis (RTA), energy- and thermal-aware scheduling, and application-specific policies. The survey identifies a critical gap between cross-platform scheduling support and formal schedulability analysis, categorizes twelve technical approaches, and highlights five open challenges—including joint optimization of timing, energy efficiency, and reliability. This work establishes the first structured, benchmark-quality survey for scheduling theory and practice in heterogeneous real-time systems.

Technology Category

• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Machine Learning: Hardware-aware ML
• Constraint Satisfaction and Optimization: Satisfiability Modulo Theories

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Web performance, measurement, and characterization
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

Accelerator-based heterogeneous architectures, such as CPU-GPU, CPU-TPU, and CPU-FPGA systems, are widely adopted to support the popular artificial intelligence (AI) algorithms that demand intensive computation. When deployed in real-time applications, such as robotics and autonomous vehicles, these architectures must meet stringent timing constraints. To summarize these achievements, this article presents a comprehensive survey of real-time scheduling techniques for accelerator-based heterogeneous platforms. It highlights key advancements from the past ten years, showcasing how proposed solutions have evolved to address the distinct challenges and requirements of these systems. This survey begins with an overview of the hardware characteristics and common task execution models used in accelerator-based heterogeneous systems. It then categorizes the reviewed works based on soft and hard deadline constraints. For soft real-time approaches, we cover real-time scheduling methods supported by hardware vendors and strategies focusing on timing-critical scheduling, energy efficiency, and thermal-aware scheduling. For hard real-time approaches, we first examine support from processor vendors. We then discuss scheduling techniques that guarantee hard deadlines (with strict response time analysis). After reviewing general soft and hard real-time scheduling methods, we explore application- or scenario-driven real-time scheduling techniques for accelerator-enabled heterogeneous computing platforms. Finally, the article concludes with a discussion of open issues and challenges within this research area.

Problem

• Survey real-time scheduling for accelerator-based heterogeneous architectures
• Address timing constraints in AI-driven time-critical applications
• Compare soft and hard deadline scheduling techniques

Innovation

• Survey of real-time scheduling techniques
• Focus on CPU-GPU, CPU-TPU, CPU-FPGA systems
• Covers soft and hard deadline constraints