Xenomai-based multiple-process system, for real-time data acquisition and graphical display control

To elicit complex and rich graphical displays, and record neuronal phenomena ofinterest while all simultaneously being capable to interact in a closed-loop with external devises is a challenging task to all neurophysiologists. To facilitate this process, we have developed an Open-Source software system using a single computer running a well established Linux architecture (Ubuntu) associated to a kernel duo providing hard real-time support (Xenomai). We show that a single computer using our API is capable, for any tasks that require OpenGL displaying, to acheive millisecond accuracy programmed events. In this report, we describe the design of our system, benchmark and its performance in a real-world setting, and describe some key features.