A Cloud Gaming System Based on NVIDIA GRID GPU

Qingdong Hou,Chu Qiu,Kaihui Mu,Quan Qi,Yongquan Lu
DOI: https://doi.org/10.1109/dcabes.2014.19
2014-01-01
Abstract:Like Internet sweeping the world 20 years ago, cloud gaming, today, is gaining in popularity with unimaginable speed. However, many technical issues must be solved through its exploring process. Among all of these problems, bandwidth limitation and support for multi-client concurrent access become the bottleneck for further development of cloud gaming. Cloud gaming is based on real-time audio and video stream, therefore, it will be a great contribution to cloud gaming on research how to encode the stream with minimum time and maximum compression radio. So we are committed to improve the ways of encoding and multi-client concurrent access in the server of cloud gaming. In order to reduce the delay in limited bandwidth and improve multi-client concurrent access efficiency, a concurrent server is built up based on the integration of NVIDIA GRID GPU and the open source cloud gaming platform Gaming Anywhere. As the results show, the latency is reduced and the concurrency is improved in this system.
What problem does this paper attempt to address?