Neuro-inspired artificial intelligence technologies for the electronics of the future
Funding by „Mikroelektronik aus Deutschland - Innovationstreiber der Digitalisierung “ des Bundesministeriums für Bildung und Forschung (BMBF)
Research-intensive microelectronics and its applications are cross-sector drivers of progress, competition and innovation. In order to accelerate the transfer of knowledge and findings to industry and to support structural change in the Rheinland region, Forschungszentrum Jülich and RWTH Aachen University aim to create an internationally visible location for top-level research in the field of memristic circuits for neuromorphic computers.
Objectives and Methodological Approach
The NEUROTEC project aims to develop innovative technologies and concepts for memristive circuits to overcome the limitations of current computer architectures, especially for AI applications. Specifically, this involves the development of artificial synapses or "Computing-in-Memory" concepts for artificial neural networks. For this purpose, different memristive materials and circuits are investigated, integration concepts for synaptic networks are developed and simulation models are created. Furthermore, characterization methods for memristive circuits and concepts for neuromorphic circuits are developed, implemented and validated.
The wide-ranging research into different memristic materials at various technological levels should enable the development of novel component technologies and computer architectures: for very powerful and at the same time particularly energy-efficient computing and artificial intelligence applications. The associated industrial partners from the region of the Rhenish mining district will be given the opportunity to establish new business areas in the field of this potentially disruptive technology at an early stage.