Answers are grounded in this paper and cite their sources.
In the past decades, video games have experienced a remarkable surge in popularity, making the gaming industry a multi-billion-dollar business. Simultaneously, game development has expanded beyond AAA publishers, with independent developers achieving impressive success. Today’s users are highly aware of quality standards and expect a reliable experience, especially in multiplayer games. However, traditional architectures, peer-to-peer and client–server, present complementary limitations, either from a user experience perspective (e.g., stability, latency, disconnection handling) or from a business perspective (e.g., deployment, maintenance, and scalability costs). In this work, we propose a novel architecture—dubbed netskip —that bridges the gap between peer-to-peer and client–server models by combining their strengths. The netskip architecture eliminates single points of failure and reduces deployment costs by removing the need for replicated servers, while still supporting disconnections, reconnections, and temporary network partitions. It manages game state in multiplayer settings using conflict-free replicated data types, and we formally prove that the netskip architecture ensures eventual consistency among nodes. Finally, we show through an empirical study using real video game traces that the netskip architecture delivers performance comparable to peer-to-peer and client–server architectures.
No takes yet. Share an insight, caveat, or question.
Bertolotti et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: