Recent Advances in Transistor‐Based Artificial Synapses

Abstract

Simulating biological synapses with electronic devices is a re-emerging field of research. It is widely recognized as the first step in hardware building brain-like computers and artificial intelligent systems. Thus far, different types of electronic devices have been proposed to mimic synaptic functions. Among them, transistor-based artificial synapses have the advantages of good stability, relatively controllable testing parameters, clear operation mechanism, and can be constructed from a variety of materials. In addition, they can perform concurrent learning, in which synaptic weight update can be performed without interrupting the signal transmission process. Synergistic control of one device can also be implemented in a transistor-based artificial synapse, which opens up the possibility of developing robust neuron networks with significantly fewer neural elements. These unique features of transistor-based artificial synapses make them more suitable for emulating synaptic functions than other types of devices. However, the development of transistor-based artificial synapses is still in its very early stages. Herein, this article presents a review of recent advances in transistor-based artificial synapses in order to give a guideline for future implementation of synaptic functions with transistors. The main challenges and research directions of transistor-based artificial synapses are also presented.In the nerve system, a synapse is a specialized structure that allows a neuron to pass chemical or electrical signals to another Figure 2. a) Schematic illustration (top) and microscopy image (bottom) of flexible synaptic transistors based on a random matrix of semiconducting CNTs. b) Case 1: the amplitudes of V LTP and V LTP are greater than other cases; thus, NL is the highest and ΔG is the largest. c) Case 2: the amplitudes of V LTP and V LTP are smaller than in case 1; thus, NL and ΔG are lower. d) Case 3: if the CNT transistor without the Au floating gate is used for the synaptic transistor, NL and ΔG are considerably smaller than in the other cases due to the limited charge storage space. Reproduced with permission. [87]

The Design of 3D‐Interface Architecture in an Ultralow‐Power, Electrospun Single‐Fiber Synaptic Transistor for Neuromorphic Computing

“…Recently, artificial intelligence and deep learning algorithms have attracted considerable attention while traditional von Neuman computers could not efficiently deal with the unstructured information because of the physical separation of the storage and processing units, as the “von Neumann bottleneck.”1–5 With the coming era of big data and artificial intelligence, the device with new architecture should be developed for comprehensive innovation of the computers. Brain‐inspired neuromorphic computation with efficient energy utilization, massive parallelism, and flexible adaptive capability, exhibits the great potential to realize multifunctional computing 6–8. There are ≈10 11 neurons in a human brain with over 1000 synaptic connections of each neuron connected with other neurons.…”

Highly Robust Flexible Ferroelectric Field Effect Transistors Operable at High Temperature with Low‐Power Consumption

“…Indeed, the output characteristics of the flexible mica/CFO/SRO/PZT/ZnO/Pt FeFET measured at room temperature shown in Figure exhibit the expected n‐type source‐drain current ( I DS ) versus source‐drain voltage ( V DS ) behavior under a series of gate voltages V GS between 0 and 6 V. It is observed that V GS tunes the carrier concentration of semiconductor channel layer and hence control the level of I DS , making it possible to access and program multiple conduction states, ideal for memristor applications such as artificial synapse, which mimics biological synapse as recently reviewed by Dai et al The I DS – V DS curves are linear for V DS biases lower than 1 V, and saturate for V DS at 6 V, indicating the flexible FeFET can operate under small voltages between 0 ≤ V DS ≤ 6 V with low power consumption, much smaller than typical flexible FeFETs based on polymers, as we will demonstrate in more details later. Furthermore, the FeFET is robust under large bending deformation with radius as small as 4 mm (Figure a), can be kept bent for 1 h without any degradation (Figure b), and retains excellent output characteristic after 500 cycles of bending test under a small radius of 6 mm (Figure c), making them excellent candidate as flexible FeFETs.…”

Recent Advances of Volatile Memristors: Devices, Mechanisms, and Applications

“…In the past years, many circuits based on CMOS technology have been used to mimic synapses. [64][65][66][67][68][69][70] However, because of the finite similarity to the biological synapses, the three terminal devices need complex circuit design to imitate the synaptic behavior. As a better choice, two terminal devices (such as memristors), with less footprint and lower power consumption, are widely utilized to mimic synapses.…”