Brain-Inspired technologies: Towards chips that think?

The advent of the Internet-of-Things has introduced a new paradigm that supports a decentralized and hierarchical communication architecture, where a great deal of analytics processing occurs at the edge and at the end-devices instead of in the Cloud. To map the embedded-systems requirements, we present a holistic research approach to the development of low-power architectures inspired by the human brain, where process development and integration, circuit design, system architecture, and learning algorithms are simultaneously optimized. This paper is organized as follows: We begin with a survey of recent research on the human brain and a historical perspective of cognitive neuroscience. Then, artificial intelligence is introduced, and the challenges of Deep Learning systems (in terms of power requirements) are addressed. The key reasons to distribute intelligence over the whole network are discussed. To emphasize the need for low-power solutions, a quantitative benchmark of existing specialized edge platforms that can execute machine-learning algorithms on conventional embedded hardware is presented. The primary focus of this paper will be on the implementation of optimized neuromorphic hardware as a highly promising solution for future ultra-low-power cognitive systems. We show that emerging technologies (such as advanced CMOS, 3D technologies, emerging resistive memories, and Silicon photonics), coupled with novel brain-inspired paradigms, such as spike-coding and spike-time-dependent-plasticity, have extraordinary potential to provide intelligent features in hardware, approaching the way knowledge is created and processed in the human brain. Finally, we conclude with our vision of the enabled future disruptive applications and a discussion of the main challenges which should be tackled to exploit the full potential of brain-inspired technologies.