Exploiting Dependency and Parallelism: Real-Time Scheduling and Analysis for GPU Tasks

TL;DR

This paper introduces a scheduling and analysis method for GPU tasks structured as DAGs, improving predictability and reducing worst-case execution times without extra hardware or software modifications.

Contribution

It presents a scalable scheduling and timing analysis approach for DAG-structured GPU tasks that reduces response time and provides safe, nonpessimistic bounds without kernel priority assumptions.

Findings

Reduces worst-case makespan by up to 32.8%

Decreases measured task execution time by up to 21.3%

Effective on synthetic and real-world benchmarks

Abstract

With the rapid advancement of Artificial Intelligence, the Graphics Processing Unit (GPU) has become increasingly essential across a growing number of safety-critical application domains. Applying a GPU is indispensable for parallel computing; however, the complex data dependencies and resource contention across kernels within a GPU task may unpredictably delay its execution time. To address these problems, this paper presents a scheduling and analysis method for Directed Acyclic Graph (DAG)-structured GPU tasks. Given a DAG representation, the proposed scheduling scales the kernel-level parallelism and establishes inter-kernel dependencies to provide a reduced and predictable DAG response time. The corresponding timing analysis yields a safe yet nonpessimistic makespan bound without any assumption on kernel priorities. The proposed method is implemented using the standard CUDA API, requiring no additional software or hardware support. Experimental results under synthetic and real-world benchmarks demonstrate that the proposed approach effectively reduces the worst-case makespan and measured task execution time compared to the existing methods up to 32.8% and 21.3%, respectively.