RESUMO
In this work, we report the digital and analog resistive-switching (RS) characteristics in a memristor based on silicon nanocrystals (Si-NCs) integrated into a complementary metal-oxide-semiconductor (MOS) structure. Si-NCs with a diameter of 5.48 ± 1.24 nm embedded in a SiO2/Si-NCs/SiO2 multilayer structure acts as an RS layer. These devices exhibit bipolar RS with an intermediate resistance step during SET and RESET processes, which is believed to lie in the Si-NCs layer acting as charge-trapping nodes. The endurance studies of about 70 DC cycles indicate an ON/OFF ratio of ~106 and a retention time larger than 104 s. Long-term potentiation (LTP, -2 V) and long-term depression (LTD, +4 V) are obtained by applying consecutive identical pulse voltages of 150 ms duration. The current value gradually increases/decreases (LTP/LTD) as the pulse number increases. Three consecutive identical pulses of -2 V/150 ms (LTP) separated by 5 and 15 min show that the last current value obtained at the end of each pulse train is kept, confirming an analog RS behavior. These characteristics provide a possible way to mimic biological synapse functions for applications in neuromorphic computing in Si-NCs-based CMOS structures.
RESUMO
Machine learning has experienced unprecedented growth in recent years, often referred to as an "artificial intelligence revolution." Biological systems inspire the fundamental approach for this new computing paradigm: using neural networks to classify large amounts of data into sorting categories. Current machine-learning schemes implement simulated neurons and synapses on standard computers based on a von Neumann architecture. This approach is inefficient in energy consumption, and thermal management, motivating the search for hardware-based systems that imitate the brain. Here, the present state of thermal management of neuromorphic computing technology and the challenges and opportunities of the energy-efficient implementation of neuromorphic devices are considered. The main features of brain-inspired computing and quantum materials for implementing neuromorphic devices are briefly described, the brain criticality and resistive switching-based neuromorphic devices are discussed, the energy and electrical considerations for spiking-based computation are presented, the fundamental features of the brain's thermal regulation are addressed, the physical mechanisms for thermal management and thermoelectric control of materials and neuromorphic devices are analyzed, and challenges and new avenues for implementing energy-efficient computing are described.
RESUMO
Liquid State Machines (LSMs) are computing reservoirs composed of recurrently connected Spiking Neural Networks which have attracted research interest for their modeling capacity of biological structures and as promising pattern recognition tools suitable for their implementation in neuromorphic processors, benefited from the modest use of computing resources in their training process. However, it has been difficult to optimize LSMs for solving complex tasks such as event-based computer vision and few implementations in large-scale neuromorphic processors have been attempted. In this work, we show that offline-trained LSMs implemented in the SpiNNaker neuromorphic processor are able to classify visual events, achieving state-of-the-art performance in the event-based N-MNIST dataset. The training of the readout layer is performed using a recent adaptation of back-propagation-through-time (BPTT) for SNNs, while the internal weights of the reservoir are kept static. Results show that mapping our LSM from a Deep Learning framework to SpiNNaker does not affect the performance of the classification task. Additionally, we show that weight quantization, which substantially reduces the memory footprint of the LSM, has a small impact on its performance.
RESUMO
A neurochip comprises a small device based on the brain-machine interfaces that emulate the functioning synapses. Its implant in the human body allows the interaction of the brain with a computer. Although the data-processing speed is still slower than that of the human brain, they are being developed. There is no ethical conflict as long as it is used for neural rehabilitation or to supply impaired or missing neurological functions. However, other applications emerge as controversial. To the best of our knowledge, there have no been publications about the neurosurgical role in the application of this neurotechnological advance. Deliberation on neurochips is primarily limited to a small circle of scholars such as neurotechnological engineers, artists, philosophers, and bioethicists. Why do we address neurosurgeons? They will be directly involved as they could be required to perform invasive procedures. Future neurosurgeons will have to be a different type of neurosurgeon. They will be part of interdisciplinary teams interacting with computer engineers, neurobiologist, and ethicists. Although a neurosurgeon is not expected to be an expert in all areas, they have to be familiar with them; they have to be prepared to determine indications, contraindications and risks of the procedures, participating in the decision-making processes, and even collaborating in the design of devices to preserve anatomic structures. Social, economic, and legal aspects are also inherent to the neurosurgical activity; therefore, these aspects should also be considered.
RESUMO
Memristive devices have found application in both random access memory and neuromorphic circuits. In particular, it is known that their behavior resembles that of neuronal synapses. However, it is not simple to come by samples of memristors and adjusting their parameters to change their response requires a laborious fabrication process. Moreover, sample to sample variability makes experimentation with memristor-based synapses even harder. The usual alternatives are to either simulate or emulate the memristive systems under study. Both methodologies require the use of accurate modeling equations. In this paper, we present a diffusive compact model of memristive behavior that has already been experimentally validated. Furthermore, we implement an emulation architecture that enables us to freely explore the synapse-like characteristics of memristors. The main advantage of emulation over simulation is that the former allows us to work with real-world circuits. Our results can give some insight into the desirable characteristics of the memristors for neuromorphic applications.
RESUMO
Neuromorphic engineering is a discipline devoted to the design and development of computational hardware that mimics the characteristics and capabilities of neuro-biological systems. In recent years, neuromorphic hardware systems have been implemented using a hybrid approach incorporating digital hardware so as to provide flexibility and scalability at the cost of power efficiency and some biological realism. This paper proposes an FPGA-based neuromorphic-like embedded system on a chip to generate locomotion patterns of periodic rhythmic movements inspired by Central Pattern Generators (CPGs). The proposed implementation follows a top-down approach where modularity and hierarchy are two desirable features. The locomotion controller is based on CPG models to produce rhythmic locomotion patterns or gaits for legged robots such as quadrupeds and hexapods. The architecture is configurable and scalable for robots with either different morphologies or different degrees of freedom (DOFs). Experiments performed on a real robot are presented and discussed. The obtained results demonstrate that the CPG-based controller provides the necessary flexibility to generate different rhythmic patterns at run-time suitable for adaptable locomotion.