Neuromorphic Computing: A Path to Artificial Intelligence Through Emulating Human Brains

Document Type

Book Chapter

Publication Date



Department of Biological Sciences; Department of Electrical and Computer Engineering; Department of Biomedical Engineering


The human brain is the most powerful computational machine in this world that has inspired artificial intelligence for many years. One of the latest outcomes of the reverse engineering neural system is deep learning, which emulates the multiple-layer structure of biological neural networks. Deep learning has achieved a variety of unprecedented successes in a large range of cognitive tasks. However, accompanied by the achievements, the shortcomings of deep learning are becoming more and more severe. These drawbacks include the demand for massive data, energy inefficiency, incomprehensibility, etc. One of the innate drawbacks of deep learning is that it implements artificial intelligence through the algorithms and software alone with no consideration of the potential limitations of computational resources. On the contrary, neuromorphic computing, also known as brain-inspired computing, emulates the biological neural networks through a software and hardware co-design approach and aims to break the shackles from the von Neumann architecture and digital representation of information within it. Thus, neuromorphic computing offers an alternative approach for next-generation AI that balances computational complexity, energy efficiency, biological plausibility, and intellectual competence. This chapter aims to comprehensively introduce neuromorphic computing from the fundamentals of biological neural systems, neuron models, to hardware implementations. Lastly, critical challenges and opportunities in neuromorphic computing are discussed.

Publication Title

Frontiers of Quality Electronic Design (QED): AI, IoT and Hardware Security