Supporting Parallelism in Server-based Multiprocessor Systems

TL;DR

This paper introduces a novel real-time scheduling approach for multiprocessor systems that combines constant bandwidth servers with work-stealing to support dynamic parallelism and improve load balancing.

Contribution

It presents a new scheduling method that ensures task isolation while allowing tasks to execute on multiple processors simultaneously.

Findings

Supports dynamic task-level parallelism in server-based systems

Ensures task isolation with combined scheduling approach

Enables tasks to run on multiple processors at once

Abstract

Developing an efficient server-based real-time scheduling solution that supports dynamic task-level parallelism is now relevant to even the desktop and embedded domains and no longer only to the high performance computing market niche. This paper proposes a novel approach that combines the constant bandwidth server abstraction with a work-stealing load balancing scheme which, while ensuring isolation among tasks, enables a task to be executed on more than one processor at a given time instant.