You are here
Following the success of the past six FeBID workshops and the Feedback Computing workshops in 2012 and 2013, the 9th International Workshop on Feedback Computing is a premier workshop in designing feedback-based systems for computing systems, including but not limited to performance and resource management, power, energy, and thermal management, and availability and reliability management. Our goal is to provide a forum for creating new ideas and agendas to broaden the theoretical foundations and applications of feedback in computing systems.
From theory’s perspective, the Feedback Computing workshop builds upon the success of the International Workshops on Feedback Control Implementation and Design in Computing Systems and Networks (FeBID) which focus primarily on classic control theory as the analytic foundation for feedback. It further expands the scope by acknowledging the need for a broader perspective on feedback in computing systems. We believe classic control-theoretical techniques represent one of several viable foundations for understanding the performance characteristics of systems that involve feedback, such as convergence, stability, and latency (among other performance metrics). Meanwhile, qualitatively different foundations are also developed for this purpose in other communities, such as machine learning, mathematical optimization, cyber-physical computing, game theory, and evolutionary graph theory. The feedback approach provides an elegant framework to incorporate these different methods and accomplish systematic management. As computing systems grow more complex to encompass clouds, social networks, distributed sensing systems, and warehouse-scale computers, new modeling, analysis, and control techniques are needed to understand, predict, and/or influence their behavior upon a feedback system. The Feedback Computing workshop will be unique in offering a single venue for exchange of ideas and insights exploring the feedback concept in computing contexts.
From application’s perspective, while being highly synergetic with the ICAC conference (another major event in the USENIX Federated Conference Week), the workshop differentiates from ICAC with its focus on the use of feedback-based design patterns in computing systems. These patterns can come from control engineers using formal mathematical models and control-theoretical techniques widely adopted in other areas of engineering. They can also come from software and system engineers with simple and workable feedback embedded heuristics in computing system design. Alternatively, they can come from game theory, graph theory, AI, or other sources that have essential feedback and dynamic system concepts implemented. The objective of this workshop is to foster collaborations between the relevant communities and between academia and industry, as well as to study and create new control patterns and integrated applications that make a real world impact.
See more information of the past workshops at the following links:
Feedback Computing 2013
June 25, 2013, San Jose, CA, USA (co-located with ICAC 2013)
Feedback Computing 2012
September 17, 2012, San Jose, CA, USA (co-located with ICAC 2012)
June 14, 2011, Karlsruhe, Germany (co-located with ICAC 2011)
April 13, 2010, Paris, France (co-located with EuroSys 2010)
April 16, 2009, San Francisco, CA, USA (co-located with CPS Week 2009)
June 6, 2008, Annapolis, MD, USA (co-located with Sigmetrics 2008)
May 25, 2007, Munich, Germany (co-located with IM 2007)
April 3, 2006, Vancouver, BC, Canada (co-located with NOMS 2006)