Accelerating Energy Games Solvers on Modern Architectures

TL;DR

This paper explores leveraging modern GPU and multi-core CPU architectures to significantly accelerate the solving of Energy Games, a type of quantitative game used in embedded controller modeling, achieving up to 36x speedup.

Contribution

It introduces four parallel implementations for Energy Games solving on modern hardware, substantially improving performance over traditional sequential methods.

Findings

Up to 36x speedup over baseline implementation

Faster convergence on real-world graphs

Effective use of GPU and multi-core CPU parallelism

Abstract

Quantitative games, where quantitative objectives are defined on weighted game arenas, provide natural tools for designing faithful models of embedded controllers. Instances of these games that recently gained interest are the so called Energy Games. The fast-known algorithm solves Energy Games in O(EVW) where W is the maximum weight. Starting from a sequential baseline implementation, we investigate the use of massively data computation capabilities supported by modern Graphics Processing Units to solve the `initial credit problem' for Energy Games. We present four different parallel implementations on multi-core CPU and GPU systems. Our solution outperforms the baseline implementation by up to 36x speedup and obtains a faster convergence time on real-world graphs.