Advance design for brain-inspired flexible computing : materials, structures, models to applications

Abstract

Recent years have witnessed a growing demand for the storage and computation of data in large-scale with the rapid development of artificial intelligence (AI). The brain-inspired computing, which are featured with many advantages, like high-speed and low power consumption, are promising candidates for meeting this demand. Inspirations have been got by learning from various aspects of human brain, with the structures, architectures, processing tactics, functions, and working manner included, to fully realize those computational advantages of the human brain. Since synapses, as one of the basic components in human brain, are featured with softness and deformability, the flexible brain-inspired computing systems have attached a lot of attention. However, reviews in the view of advanced design for the flexible brain-inspired computing are few. Here, the newest development is comprehensively reviewed in this field. First, background and mechanisms are briefed, and then most important considerations for material, structure and model designs are illustrated together with solutions to realize the device flexibilities and high-level brain intelligence. Next, strategies are summarized to obtain brain-inspired computing with high overall performance. After that, the design thoughts for future flexible brain-inspired computing are proposed. Finally, some perspectives are put forward.