Synaptic Devices and Artificial Neurons for Neuromorphic Computation

Dear Colleagues,

Using neuron and synapses-mimetic devices, neuromorphic computation combines memory with computing in the same unit and thus has much higher efficiency in processing complex information than the current von Neumann architecture-based computers. The neuromorphic hardware systems can be divided into deep neural networks (DNNs) and spiking neural networks (SNNs). The DNN system involves artificial synaptic devices with multilevel modulated conductance being adjusted to desired values in the backpropagation training, which serve as the synaptic weights. During information processing, the input vector from the presynaptic neuron array is multiplied by the whole synaptic weight matrix while transferring the signals to the post-synaptic neuron array. To achieve high accuracy, the synaptic device in DNNs should have a linear conductance variation with more than 100 states. In contrast, an SNN system behaves more like the human brain by using spike signaling and learning algorithms such as spike-timing-dependent plasticity (STDP) for unsupervised learning, and hence it requires synaptic devices with STDP characteristics, which have been less studied than their counterpart in DNNs. In addition to artificial synapses, quality devices with integrate-and-fire characteristics to simulate biological neurons are also required in both DNNs and SNNs, even though the number of artificial neurons are much less than that of synapses. Such an artificial neuron integrates input currents through synapses and fires spikes to the next synaptic layer if the integrated effect from the input currents exceeds a threshold value, which is decided by the particular material property.

There has been a huge interest in developing non-Si based two- or three-terminal devices as artificial neurons and synapses. Various materials and devices such as memristors or resistive memory, phase change memory, electrochemical transistors, ferroelectric transistors, and charge trapping transistors have been studied as artificial synaptic devices, while phase change materials, ferroelectric materials, and others have also been evaluated as artificial neurons. It is expected that these non-Si based artificial synapses and neurons will eventually be integrated into more compact DNN or SNN systems for efficient neuromorphic computing.

The aim of this special Issue is to report the recent advances in novel material-based artificial synaptic or neuron devices. We are confident that the publication of such a Special Issue will stimulate the curiosity of researchers to explore novel materials and structures to further advance the progress in this promising field and eventually put forward their practical applications in artificial intelligence.

It is our pleasure to invite you to submit a manuscript reporting new materials and structures, fundamental mechanisms, novel device concepts, device performances, as well as other related topics for this Special Issue. Full papers, communications, and reviews are all welcome.

Prof. Dr. Zhaoyang Fan
Guest Editor

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Electronics is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2400 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

• neuromorphic computing hardware
• artificial synapses
• artificial neurons
• synaptic devices
• artificial neuron networks