This review provides a comprehensive analysis of artificial neurons starting with the description of biological neurons and discusses the pathway for enabling this brain‐inspired technology by evaluating system‐level architectures and deriving application‐specific requirements for real‐world deployment.
Abstract
The fundamental energy and latency limitations of the von Neumann architecture have necessitated a paradigm shift toward neuromorphic systems. This approach addresses the “memory wall” bottleneck by emulating the brain's event‐driven and massively parallel processing capabilities. However, the physical realization of these neurons in hardware remains a critical challenge. While complementary metal–oxide–semiconductor (CMOS) based implementations can simulate neuronal dynamics, they are constrained by high transistor counts and excessive power consumption, which limits their scalability for high‐density integration. This review provides a comprehensive analysis of artificial neurons starting with the description of biological neurons. The various neuron models and learning models are discussed to establish the functional requirements for hardware emulation. We subsequently detail the limitations of CMOS neurons as a motivation to the transition toward post‐CMOS neurons. These include emerging device‐based implementations of integrate‐and‐fire neurons, Hodgkin–Huxley neurons, and other neurons with novel functionalities. Furthermore, we discuss the pathway for enabling this brain‐inspired technology by evaluating system‐level architectures and deriving application‐specific requirements for real‐world deployment. Finally, we present conclusions and future outlooks for achieving energy‐efficient and high‐density neuromorphic computing systems.
The energy demand of modern artificial intelligence (AI) has grown faster than the efficiency of the von Neumann hardware that runs it, renewing interest in neuromorphic computing as a brain-inspired alternative. This review surveys more than three decades of neuromorphic hardware, from Mead's foundational analog very-...
R. Alajangi· Ci-STEM Journal of Digital T...· 0 citations
The rapid growth of data‐intensive computing has exposed the physical limitations of conventional hardware architectures. While accumulative multistate memory devices offer low‐power alternatives, incorporating multistate reconfigurability into switching devices enables fully integrated, multifunctional neuromorphic...
Dongmi Heo, Min-Hu Huang, Youngseok Kim et al.· Advanced Functional Material...· 0 citations
This chapter explores the critical role of emerging non-volatile memory (eNVM) devices in advancing neuromorphic computing. Inspired by the human brain's remarkable ability to perform complex tasks like pattern recognition and adaptation, neuromorphic computing offers immense potential for implementing complex neural n...
Brain-like synaptic electronics represent a promising avenue for next-generation information technology, addressing the limitations of conventional von Neumann architectures. Compared with existing CMOS-based neuromorphic hardware, highly scalable artificial synaptic device units offer superior computational energy eff...
Guang-Yue Shen, Heng Liang, Han Tang et al.· Journal of Physics D: Applie...· 0 citations
Memristor‐based neuromorphic computing has emerged as a promising hardware paradigm for energy‐efficient, scalable artificial intelligence edge applications. This review surveys recent progress in memristor‐based neuromorphic computing, in‐memory computing architectures, learning algorithms, and hardware implementati...
Synaptic transistors (STs) are gaining valuable interest as they hold a significant potential for neuromorphic hardware implementation. They are key elements in emulating biological synaptic behavior as they have the ability to perform memory storage and parallel information processing. Their applications extend beyond...
Z. Hafdi· Engineering Research Express· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.