Next-generation non-volatile memories with ultrafast speed, low power consumption, and high density are highly desired in the era of big data. Here, we report a high performance memristor based on a Ag/BaTiO 3 /Nb:SrTiO 3 ferroelectric tunnel junction (FTJ) with the fastest operation speed (600 ps) and the highest number of states (32 states or 5 bits) per cell among the reported FTJs. The sub-nanosecond resistive switching maintains up to 358 K, and the write current density is as low as 4 × 10 3 A cm −2. The functionality of spike-timingdependent plasticity served as a solid synaptic device is also obtained with ultrafast operation. Furthermore, it is demonstrated that a Nb:SrTiO 3 electrode with a higher carrier concentration and a metal electrode with lower work function tend to improve the operation speed. These results may throw light on the way for overcoming the storage performance gap between different levels of the memory hierarchy and developing ultrafast neuromorphic computing systems.

The rapid development of neuro-inspired computing demands synaptic devices with ultrafast speed, low power consumption, and multiple non-volatile states, among other features. Here, a high-performance synaptic device is designed and established based on a Ag/PbZr0.52Ti0.48O3 (PZT, (111)-oriented)/Nb:SrTiO3 ferroelectric tunnel junction (FTJ). The advantages of (111)-oriented PZT (~1.2 nm) include its multiple ferroelectric switching dynamics, ultrafine ferroelectric domains, and small coercive voltage. The FTJ shows high-precision (256 states, 8 bits), reproducible (cycle-to-cycle variation, ~2.06%), linear (nonlinearity <1) and symmetric weight updates, with a good endurance of >109 cycles and an ultralow write energy consumption. In particular, manipulations among 150 states are realized under subnanosecond (~630 ps) pulse voltages ≤5 V, and the fastest resistance switching at 300 ps for the FTJs is achieved by voltages <13 V. Based on the experimental performance, the convolutional neural network simulation achieves a high online learning accuracy of ~94.7% for recognizing fashion product images, close to the calculated result of ~95.6% by floating-point-based convolutional neural network software. Interestingly, the FTJ-based neural network is very robust to input image noise, showing potential for practical applications. This work represents an important improvement in FTJs towards building neuro-inspired computing systems.

Flexible ferroelectric devices have been a hot-spot topic because of their potential wearable applications as nonvolatile memories and sensors. Here, high-quality (111)-oriented BiFeO 3 ferroelectric films are grown on flexible mica substrates through an appropriate design of SrRuO 3 /BaTiO 3 double buffer layers. BiFeO 3 exhibits the largest polarization (saturated polarization P s ≈ 100 μC/cm 2 , remnant polarization P r ≈ 97 μC/cm 2 ) among all the reported flexible ferroelectric films, and ferroelectric polarization is very stable in 10 4 bending cycles under 5 mm radius. Accordingly, the ferroelectric memristor behaviors are demonstrated with continuously tunable resistances, and thus, the functionality of spike-timing-dependent plasticity is achieved, indicating the capability of flexible BiFeO 3 -based memristors as solid synaptic devices. Moreover, in artificial neural network simulations based on the experimental characteristics of the memristor, a high recognition accuracy of ∼90% on handwritten digits is obtained through online supervised learning. These results highlight the potential wearable applications of flexible ferroelectric memristors for data storage and computing.

Ferroic‐order‐based devices are emerging as alternatives to high density, high switching speed, and low‐power memories. Here, multi‐nonvolatile resistive states with a switching speed of 6 ns and a write current density of about 3 × 103 A cm−2 are demonstrated in crossbar‐structured memories based on all‐oxide La0.7Sr0.3MnO3/BaTiO3/La0.7Sr0.3MnO3 multiferroic tunnel junctions. The tunneling resistive switching as a function of voltage pulse duration time, associated with the ferroelectric domain reversal dynamics, is ruled by the Kolmogorov–Avrami–Ishibashi switching model with a Lorentzian distribution of characteristic switching time. It is found that the characteristic resistance switching time decreases with increasing voltage pulse amplitude following Merz's law and the estimated write speed can be less than 6 ns at a relatively higher voltage. These findings highlight the potential application of multiferroic devices in high speed, low power, and high‐density memories.

No abstract