The research group “Cognitronics and Sensor Systems”, lead by Professor Ulrich Rückert, focuses on massively parallel and reconfigurable system architectures. Our area of interest covers embedded processor architectures and energy-efficient sensor nodes as well as scalable server platforms. On the application level, we emphasize distributed cognitive systems characterized by a decentralized organization utilizing architectures that autonomously adapt their system resources to changing environmental conditions or processing requirements.
Within eProcessor, we utilize our expertise in resource-efficient system architectures for HPC and cloud computing, system design, reconfigurable computing, machine learning, and algorithms and applications for Edge computing. We focus on bring-up, test, and integration of the eProcessor AISC into the overall hardware platform. Using an industry-standard microserver form factor such as COM Express or COM-HPC allows easy integration into third-party hardware platforms and applications, facilitating broader evaluation and use of the eProcessor ASIC. Additionally, we are active in porting a Smart Mirror application towards the eProcessor platform. The Smart Mirror integrates multiple AI-based software packages such as image recognition (i.e., face, gesture, and object), sound identification, and behavior prediction. eProcessor utilizes hardware acceleration to substantially improve both the performance and energy efficiency of the Smart Mirror application.