Handling Parallelism in a Concurrency Model

TL;DR

This paper introduces a library-based method to enable safe parallel array access in a concurrency model, specifically SCOOP, without significant performance loss, thus expanding its applicability to data-parallel applications.

Contribution

The paper presents a novel library approach that allows parallel data access in SCOOP, maintaining safety guarantees and minimal performance overhead.

Findings

Negligible performance overhead in parallel benchmarks

Effective preservation of safety guarantees

Enhanced applicability of SCOOP for data parallelism

Abstract

Programming models for concurrency are optimized for dealing with nondeterminism, for example to handle asynchronously arriving events. To shield the developer from data race errors effectively, such models may prevent shared access to data altogether. However, this restriction also makes them unsuitable for applications that require data parallelism. We present a library-based approach for permitting parallel access to arrays while preserving the safety guarantees of the original model. When applied to SCOOP, an object-oriented concurrency model, the approach exhibits a negligible performance overhead compared to ordinary threaded implementations of two parallel benchmark programs.