Abstract
There is an emerging class of real‐time interactive applications that require the dynamic integration of task and data parallelism. An example is the Smart Kiosk, a free‐standing computer device that provides information and entertainment to people in public spaces. The kiosk interface is computationally demanding: It employs vision and speech sensing and an animated graphical talking face for output. The computational demands of an interactive kiosk can vary widely with the number of customers and the state of the interaction. Unfortunately this makes it difficult to apply current techniques for integrated task and data parallel computing, which can produce optimal decompositions for static problems. Using experimental results from a color‐based people tracking module, we demonstrate the existence of a small number of distinct operating regimes in the kiosk application. We refer to this type of program behavior asconstrained dynamism. An application exhibiting constrained dynamism can execute efficiently by dynamically switching among a small number of statically determined fixed data parallel strategies. We present a novel framework for integrating task and data parallelism for applications that exhibit constrained dynamism. Our solution has been implemented usingStampede, a cluster programming system developed at the Cambridge Research Laboratory.
Highlights
There is an emerging class of real-time interactive applications that require dynamic integration of task and data parallelism for effective computation
We describe a novel approach to the dynamic integration of task and data parallelism for interactive real-time applications like the Cambridge Research Laboratory (CRL) Smart Kiosk
We present a general framework for integrating data parallelism into a task parallel substrate such as Stampede
Summary
There is an emerging class of real-time interactive applications that require dynamic integration of task and data parallelism for effective computation. The computational requirements of even simple vision algorithms for the kiosk are proportional to the number of customers, which cannot be predicted in advance We describe a novel approach to the dynamic integration of task and data parallelism for interactive real-time applications like the CRL Smart Kiosk. They contain our framework for dynamic integration of task and data parallelism.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have