Abstract

This thesis focuses on enhancing the scalability and memory coherence of Monte Carlo path-tracing algorithms for multi-user applications on GPUs. The research is divided into two key objectives: first, modifying the Monte Carlo path-tracing algorithm to improve its scalability for multiple users; and second, improving the memory coherence of the proposed algorithm on GPUs. Experimental results demonstrated that the proposed algorithm scales in terms of required computing resources based on scene size rather than the number of users, offering a significant improvement over traditional cloud gaming approaches, where a separate rendering instance is used for each individual user.

Resources

Thesis
Code

BibTex reference

@misc{ vandersanden2023scalable, 
        author = {Vandersanden, Jente},
        title = {A Scalable and Coherent Approach to Monte Carlo Path Tracing for Multi-User Environments},
        year = {2023}
}