Loading

• EVENTS
• Upcoming
• Calendar
• Colloquia
• TCSDLS
• Student Talks
• Research Seminars
• inDuke Seminars
• Search
• Archive
• 2013-2014 2012-2013 2011-2012 2010-2011 2009-2010 2008-2009 2007-2008 2006-2007 2005-2006 2004-2005 2003-2004 2002-2003 2001-2002 2000-2001 1999-2000 1998-1999 1997-1998 1996-1997 1995-1996 1994-1995

There has been a surge in cloud-gaming offerings where users can play video games with high-quality graphics without requiring powerful and expensive hardware. The approach taken by commercial cloud-gaming vendors such as OnLive is very faithful to the thin-client computing model: the video game is rendered completely on a cloud-server, and the visual output is streamed to the client as a H.264-compressed video.

Eduardo Cuervo, in his PhD dissertation, proposed Kahawai, an alternative approach to cloud-gaming. In Kahawai, the client is not treated simply as a dumb-terminal. Instead, it performs part of the rendering in conjunction with the cloud-server, and the results from both computations are finally used to create the high-quality output at the client. This technique significantly reduces the amount of network bandwidth used compared to the pure thin-client approach.

In my RIP, I will extend Kahawai to networked multiplayer games. In general, the playability of such games is most affected by latency: the amount of time taken for the user to receive a response to her in-game actions. The higher the latency, the more unplayable the game becomes. In a cloud-gaming setting, the cloud-server adds latencies to games, even if they are single-player games. Hence, the main challenge in my RIP is in minimizing latencies to enable enjoyable multiplayer games in Kahawai.

Advisor(s): Landon Cox

Jeffrey Chase, Bruce Maggs