Bio: Daniele Ielmini is an academic researcher from Polytechnic University of Milan. The author has contributed to research in topic(s): Resistive random-access memory & Phase-change memory. The author has an hindex of 68, co-authored 367 publication(s) receiving 16443 citation(s). Previous affiliations of Daniele Ielmini include Instituto Politécnico Nacional & Corecom.
Topics: Resistive random-access memory, Phase-change memory, Non-volatile memory, Neuromorphic engineering, Artificial neural network
Abstract: This work has been supported by the generous Baseline funding program of the King Abdullah University of Science and Technology (KAUST).
Abstract: Nowadays, artificial neural networks (ANNs) can outperform the human brain’s ability in specific tasks. However, ANNs cannot replicate the efficient and low-power learning, adaptation, and consolidation typical of biological organisms. Here, we present a hardware design based on arrays of SiO x resistive switching random-access memory (RRAM), which allows combining the accuracy of convolutional neural networks with the flexibility of bio-inspired neuronal plasticity. In order to enable the combination of the stable and the plastic attributes of the network, we exploit the spike-frequency adaptation of the neurons relying on the multilevel programming of the RRAM devices. This procedure enhances the efficiency and accuracy of the network for MNIST, noisy MNIST (N-MNIST), Fashion-MNIST, and CIFAR-10 datasets, with inference accuracies of about 99%–89%, respectively. We also demonstrate that the hardware is capable of asynchronous self-adaptation of its operative frequency according to the fire rate of the spiking neuron, thus optimizing the whole behavior of the network. We finally show that the system enables fast and accurate filter retraining to overcome catastrophic forgetting, showing high efficiency in terms of operations per second and robustness against device non-idealities. This work paves the way for the theoretical modeling and hardware realization of resilient autonomous systems in dynamic environments.
Abstract: Neuromorphic computing aims at the realization of intelligent systems able to process information similarly to our brain. Brain-inspired computing paradigms have been implemented in crossbar arrays of memristive devices; however, this approach does not emulate the topology and the emergent behaviour of biological neuronal circuits, where the principle of self-organization regulates both structure and function. Here, we report on in materia reservoir computing in a fully memristive architecture based on self-organized nanowire networks. Thanks to the functional synaptic connectivity with nonlinear dynamics and fading memory properties, the designless nanowire complex network acts as a network-wide physical reservoir able to map spatio-temporal inputs into a feature space that can be analysed by a memristive resistive switching memory read-out layer. Computing capabilities, including recognition of spatio-temporal patterns and time-series prediction, show that the emergent memristive behaviour of nanowire networks allows in materia implementation of brain-inspired computing paradigms characterized by a reduced training cost.
Abstract: Understanding the switching mechanism of the volatile resistive switching random access memory (RRAM) device is important to harness its characteristics and further enhance its performance. Accurate modeling of its dynamic behavior is also of deep value for its applications both as selector and as short-term memory synapse for future neuromorphic applications operating in temporal domain. In this work, we investigate the switching and retention (relaxation) processes of the Ag-based metallic filamentary volatile resistive switching devices. We find that the switching process can be modeled by the ionic drift under electric field, while the retention process can be modeled by the ionic diffusion along the filament surface driven by the gradient of surface atomic concentration. Through further theoretical analysis, we also find that the ionic drift and ionic diffusion can be unified within the general Einstein relation. To confirm this relation, we collect ionic mobility and diffusivity data from the literature using the switching and retention model. Finally, we show that the read voltage dependent retention time can be explained by the competition between the ionic drift and diffusion flux.
Abstract: The rapid development of big-data analytics (BDA), internet of things (IoT) and artificial intelligent Technology (AI) demand outstanding electronic devices and systems with faster processing speed, lower power consumption, and smarter computer architecture. Memristor, as a promising Non-Volatile Memory (NVM) device, can effectively mimic biological synapse, and has been widely studied in recent years. The appearance and development of two-dimensional materials (2D material) accelerate and boost the progress of memristor systems owing to a bunch of the particularity of 2D material compared to conventional transition metal oxides (TMOs), therefore, 2D material-based memristors are called as new-generation intelligent memristors. In this review, the memristive (resistive switching) phenomena and the development of new-generation memristors are demonstrated involving graphene (GR), transition-metal dichalcogenides (TMDs) and hexagonal boron nitride (h-BN) based memristors. Moreover, the related progress of memristive mechanisms is remarked.
Abstract: Amorphous to crystalline phase transition in Ge2Sb2Te5 film under the influence of a single femtosecond laser pulse is studied. Two-dimensional temperature calculations and kinetic model for crystallization were used to support experimental results and then to explain the fast mechanism of crystallization. Based on comparison with the experimental data, the theoretical Time-Temperature-Transformation diagram was calculated, that allowed to define the range of cooling rates at which crystallization is possible. The distribution of crystalline fraction in the thin film was calculated using these rates. Reflectance of the simulated structures turned out to be in good agreement with experimental observations.
Abstract: Phase-change materials are among the leading candidates that satisfy the need for in-memory computing or computational memory. Ge2Sb2Te5 (GST) is the representative material among them, but amorphous resistance drift in GST limits its cyclable stability. Here, we optimize the properties and microstructure of a GST material by introducing a BiSb phase. Results reveal that the amorphous resistance of GST decreases after the addition of BiSb and the drift coefficient can be reduced to 0.004 because of the decreased disorder in the amorphous state. Further analysis of microstructural characteristics reveals that the distribution of the Bi-Sb phase restrains the diffusion of Sb/Te element and suppresses the formation of voids in GST phases. These features stabilize the nanostructures and suppress resistance drift to a certain extent because of the dual-phase coexistence. Moreover, the conduction of GST compounds changes from a p-type to an n-type, originating from n-type BiSb precipitations. Notably, resistance drift can be further decreased through liquid nitrogen treatment, which restricts atomic mobility in small volumes, changes electron binding energy, and improves the stability of amorphous phases during structural relaxation. Decrease in resistance drift originates from the narrowing of the band gap from 0.65 eV for GST to 0.33 eV for (BiSb)5.1(GST)94.9. Thus, our present research shows that doping GST with BiSb is one of the effective means to obtain ultralow resistance shift in phase-change neuron synaptic devices.
Amra Peles1•Institutions (1)
Abstract: Determining microscopic origin of electronic conductivity in strontium ferrites is a pivotal step toward accomplishing control of a reversible topotactic phase transformation. This functionality is essential for use of these oxides in resistive switching non-volatile memory devices and future neuromorphic computing. Here, we use first principles computations for in-depth study of native defects and their effect on electronic structure in SrFeO2.5. We elucidate the role native defects play and show that an uptake of oxygen in diluted concentrations leads to a p-type conductivity in the SrFeO2.5. We identify two acceptor defect transition states at 0.25 eV for atomic and 0.1 eV for molecular oxygen interstitials.
Author's H-index: 68