Memristors integrated into a crossbar‐array architecture (CAA) are promising candidates for nonvolatile memory elements in artificial neural networks. However, the relatively low reliability of memristors coupled with crosstalk and sneak currents in CAAs have limited the realization of the full potential of this technology. Here, high‐reliability Na‐doped TiO2 memristors grown in situ by atomic layer deposition (ALD) are demonstrated, where reversible Na migration underlies the resistive‐switching mechanism. By employing ALD growth with an aqueous NaOH reactant in deionized water, uniform implantation of Na dopants is achieved in the crystallized TiO2 thin films at 250 °C without post‐annealing. The resulting Na‐doped TiO2 memristors show electroforming‐free and self‐rectifying resistive‐switching behavior, and they are ideally suited for selectorless CAAs. Effective addressing of selectorless nodes is demonstrated via electrical measurement of individual memristors in a 6 × 6 crossbar using a read current of less than 1 µA with negligible sneak current at or below the noise level of ≈100 pA. Finally, the long‐term potentiation and depression synaptic behavior from these Na‐doped TiO2 memristors achieves greater than 99.1% accuracy for image‐recognition tasks using a convolutional neural network based on the selectorless of crossbar arrays.

Artificial intelligence and machine learning are growing computing paradigms, but current algorithms incur undesirable energy costs on conventional hardware platforms, thus motivating the exploration of more efficient neuromorphic architectures. Toward this end, we introduce here a memtransistor with gate-tunable dynamic learning behavior. By fabricating memtransistors from monolayer MoS 2 grown on sapphire, the relative importance of the vertical field effect from the gate is enhanced, thereby heightening reconfigurability of the device response. Inspired by biological systems, gate pulses are used to modulate potentiation and depression, resulting in diverse learning curves and simplified spike-timing-dependent plasticity that facilitate unsupervised learning in simulated spiking neural networks. This capability also enables continuous learning, which is a previously underexplored cognitive concept in neuromorphic computing. Overall, this work demonstrates that the reconfigurability of memtransistors provides unique hardware accelerator opportunities for energy efficient artificial intelligence and machine learning.

Emerging energy-efficient neuromorphic circuits are based on hardware implementation of artificial neural networks (ANNs) that employ the biomimetic functions of memristors. Specifically, crossbar array memristive architectures are able to perform ANN vector-matrix multiplication more efficiently than conventional CMOS hardware. Memristors with specific characteristics, such as ohmic behavior in all resistance states in addition to symmetric and linear long-term potentiation/depression (LTP/LTD), are required in order to fully realize these benefits. Here, we demonstrate a Li-based composite memristor (LCM) that achieves these objectives. The LCM consists of three phases: Li-doped TiO 2 as a Li reservoir, Li 4 Ti 5 O 12 as the insulating phase, and Li 7 Ti 5 O 12 as the metallic phase, where resistive switching correlates with the change in the relative fraction of the metallic and insulating phases. The LCM exhibits a symmetric and gradual resistive switching behavior for both set and reset operations during a full bias sweep cycle. This symmetric and linear weight update is uniquely enabled by the symmetric bidirectional migration of Li ions, which leads to gradual changes in the relative fraction of the metallic phase in the film. The optimized LCM in ANN simulation showed that exceptionally high accuracy in image classification is realized in fewer training steps compared to the nonlinear behavior of conventional memristors.

No abstract

No abstract