FIGURE - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
A) traditional Von-Neumann architecture (B) Brain-inspired architecture (Burr et al., ,,,,; Silver et al., ,,,,).
Source publication
As information technology is moving toward the era of big data, the traditional Von-Neumann architecture shows limitations in performance. The field of computing has already struggled with the latency and bandwidth required to access memory (“the memory wall”) and energy dissipation (“the power wall”). These challenging issues, such as “the memory...
Similar publications
The data shuttling between computing and memory dominates the power consumption and time delay in electronic computing systems due to the bottleneck of the von Neumann architecture. To increase computational efficiency and reduce power consumption, photonic in-memory computing architecture based on phase change material (PCM) is attracting increasi...
Citations
... As a result, there is a pressing need to explore innovative computing approaches that can keep up with the demands of artificial intelligence in today's fast-paced society [1][2][3][4][5]. Motivated by the cognitive abilities of the human brain, such as perception, memory, learning, and efficient computing behavior, scientists are now directing their attention toward creating novel devices that mimic the information processing mechanisms of our brains [6][7][8][9]. In contrast to the conventional approach of neuromorphic computation, the emerging memristor device enables effective parallel computations by leveraging Ohm's law or Kirchhoff's law. ...
As a potential solution for neuromorphic applications, memristor is widely considered as a highly promising option to replicate biological synapses owing to its distinctive analog characteristics and diverse plasticity. In this study, we propose using Ag-sericin composite memristors to mimic the biological synaptic function. The memristor with a sandwich structure, consisting of Ag-sericin composite as the functional layer was fabricated. Ag-sericin memrister demonstrates typical analog resistive switching characteristics and exhibits excellent conductance modulation capability. The resistive switching characteristics are primarily determined by the formation and rupture of silver conductive paths within the sericin matrix. The synaptic plasticities, such as paired pulse facilitation, spike time-dependent plasticity, spike amplitude-dependent plasticity, spike frequency-dependent plasticity, and potentiation and depression, have been successfully replicated using Ag-sericin memristers. These findings will contribute to the advancement of intelligent computing and bionics.
... Device endurance refers to the ability of the neuron device to withstand prolonged operation without significant degradation or performance deterioration. As discussed in [24], the durability of a neuron can be extended by selecting the appropriate material for device fabrication. It can also be extended by reducing Si-HfO 2 interface traps which occur due to the immature process technology for growing high-κ dielectrics over silicon [25]. ...
In view of the soaring demand of highly scalable and energy-efficient neuron devices for future neuromorphic computing, this work demonstrates a novel double-gate ferroelectric junctionless field effect transistor (DG-FE-JLFET) neuron in 20 nm technology node. The proposed device has a homogenously doped structure with GaAs as channel material and HfZrO2 as gate oxide. Using the calibrated simulations in Atlas TCAD, it is confirmed that the proposed neuron accurately mimics the spiking behavior of the biological neuron with an energy consumption of 24 aJ/spike, which is ~ 105 folds lesser than the previously proposed single transistor neurons. The proposed design does not need additional circuitry to operate. This greatly simplifies design complexity and can also achieve higher neuron density which is important for designing large scale integration neuromorphic chips. In contrast to the previously reported JLFET neurons that work on impact ionization phenomenon, DG-FE-JLFET works on tunnelling mechanism that eradicates the need to create virtual potential well to accommodate charge carriers in the body. Finally, to validate the practical applicability, the proposed neuron has been explored for image classification with an accuracy of 93.28%.
Designing next‐generation molecular devices typically necessitates plentiful oxygen‐bearing sites to facilitate multiple‐electron transfers. However, the theoretical limits of existing materials for energy conversion and information storage devices make it inevitable to hunt for new competitors. Polyoxometalates (POMs), a unique class of metal‐oxide clusters, have been investigated exponentially due to their structural diversity and tunable redox properties. POMs behave as electron‐sponges owing to their intrinsic ability of reversible uptake‐release of multiple electrons. In this review, numerous POM‐frameworks together with desired features of a contender material and inherited properties of POMs are systematically discussed to demonstrate how and why the electron‐sponge‐like nature of POMs is beneficial to design next‐generation water oxidation/reduction electrocatalysts, and neuromorphic nonvolatile resistance‐switching random‐access memory devices. The aim is to converge the attention of scientists who are working separately on electrocatalysts and memory devices, on a point that, although the application types are different, they all hunt for a material that could exhibit electron‐sponge‐like feature to realize boosted performances and thus, encouraging the scientists of two completely different fields to explore POMs as imperious contenders to design next‐generation nanodevices. Finally, challenges and promising prospects in this research field are also highlighted.
The potential of neuromorphic computing to bring about revolutionary advancements in multiple disciplines, such as artificial intelligence (AI), robotics, neurology, and cognitive science, is well recognised. This paper presents a comprehensive survey of current advancements in the use of machine learning techniques for the logical development of neuromorphic materials for engineering solutions. The amalgamation of neuromorphic technology and material design possesses the potential to fundamentally revolutionise the procedure of material exploration, optimise material architectures at the atomic or molecular level, foster self-adaptive materials, augment energy efficiency, and enhance the efficacy of brain–machine interfaces (BMIs). Consequently, it has the potential to bring about a paradigm shift in various sectors and generate innovative prospects within the fields of material science and engineering. The objective of this study is to advance the field of artificial intelligence (AI) by creating hardware for neural networks that is energy-efficient. Additionally, the research attempts to improve neuron models, learning algorithms, and learning rules. The ultimate goal is to bring about a transformative impact on AI and better the overall efficiency of computer systems.
Autograft replaced by a nerve guidance conduit (NGC) is challenging in peripheral nerve injury because current NGC is still limited by precise conductivity and excellent biocompatibility in vivo, which influences the peripheral nerve repair even for a long lesion gap repair. Several particular elements have the potential function for nerve conductivity acceleration based on the traditional three factors of neural tissue engineering. The review aims to address three questions: 1) What is the superior factor for nerve conduction in the application? 2) How can a more conductive regenerative scaffold be constructed in vivo? 3) What is the next step in nerve regeneration for NGC? The bibliometrics analysis of NGC‐related references is adopted to acquire that the conductive material, manufacturing technology of neural scaffold, and electrical stimulation (ES) play essential roles in the acceleration of nerve conduction. This review visually analyses the research status and summarizes the main types of conductive materials, the manufacturing technologies of neural scaffolds, and the characteristics of ES. The viewpoints and outlook of developing NGC are also discussed in this review. The proposed three elements are expected to improve the nerve conduction of NGC in vivo and even address the dilemma of long‐distance peripheral nerve injury.
Binary metal oxide memristor-based artificial neural synapses offer the advantages of high-density information storage, high-efficiency computing, low power consumption and strong antiradiation damage and are therefore promising aerospace electronic devices. However, the unclear damage trends and mechanisms causing nonlinearity and asymmetry during conductance tuning under irradiation have restricted further development. Here, the damage trends and mechanisms caused by proton irradiation in Au/HfO
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> x </sub>
/TiO
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> x </sub>
/Ti memristors were studied. Results show that when the fluence of proton (25 MeV) irradiation increases to 2×10
<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">11</sup>
p/cm
<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>
, the Au/HfO
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> x </sub>
/TiO
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> x </sub>
/Ti memristor cannot return to its original high-resistance state upon conductance reduction, making the conductance tuning less linear and symmetric. X-ray photoelectron spectroscopy analysis revealed that the mechanism underlying the decrease in the linearity and symmetry of the conductance tuning is proton irradiation increasing oxygen vacancies in the heterojunction. These results can guide better design of metal oxide memristive devices suitable for aerospace devices and better modeling of devices with nonideal properties under irradiation.
