Benchmarking Performance of Unity’s Data Oriented Technology Stack
Document Type
Conference Proceeding
Publication Date
4-15-2025
Abstract
Previous research demonstrated the implementation of a simple flight dynamic model (FDM) using Unity’s Data Oriented Technology Stack (DOTS) package [4]. Although the research provided a brief demonstration and analysis of the core features of the package, most notably the entity component system (ECS) pattern, it did not exhaust all of the means by which the architectural pattern is most often used to leverage performance benefits. In this research, we extend the functionality of the flight simulator to highlight the advantages of a data-oriented programming approach over the traditional object-oriented paradigm. Features such as flight controls and environment assets are implemented as entity component systems and the multi-threading capabilities of the DOTS package is exploited by spawning and manipulating multiple aircraft. In addition, we compare the performance metrics of the data-oriented design to a similar simulation created using the legacy object-oriented paradigm. The results indicate that the data-oriented simulation performs better than the object-oriented model for CPU performance, frames per second (FPS), and memory usage.
Source Publication
CCIS Volume 2723: AI Revolution: Research, Ethics and Society (AIR-RES 2025)
Recommended Citation
Martin, B., Visci, J., Visci, M., Thompson, J., Hodson, D. (2026). Benchmarking Performance of Unity’s Data Oriented Technology Stack. In: Arabnia, H.R., Deligiannidis, L., Amirian, S., Ghareh Mohammadi, F., Shenavarmasouleh, F. (eds) AI Revolution: Research, Ethics and Society. AIR-RES 2025. Communications in Computer and Information Science, vol 2723. Springer, Cham. https://doi.org/10.1007/978-3-032-13056-3_40
Comments
© The Authors, under exclusive license to Springer Nature Switzerland AG 2026.
This conference paper was presented at AI-RES 25 in April 2025.
The full paper is accessible from Springer by subscription or purchase using the DOI link below.
Co-author B. Martin was an AFIT graduate student at the time of the conference.