To implement artificial neural networks (ANNs) based on memristor devices, it is essential to secure the linearity and symmetry in weight update characteristics of the memristor, and reliability in the cycle-to-cycle and device-to-device variations. This study experimentally demonstrated and compared the filamentary and interface-type resistive switching (RS) behaviors of tantalum oxide (Ta2O5 and TaO2)-based devices grown by atomic layer deposition (ALD) to propose a suitable RS type in terms of reliability and weight update characteristics. Although Ta2O5 is a strong candidate for memristor, the filament-type RS behavior of Ta2O5 does not fit well with ANNs demanding analog memory characteristics. Therefore, this study newly designed an interface-type TaO2 memristor and compared it to a filament type of Ta2O5 memristor to secure the weight update characteristics and reliability. The TaO2-based interface-type memristor exhibited gradual RS characteristics and area dependency in both high- and low-resistance states. In addition, compared to the filamentary memristor, the RS behaviors of the TaO2-based interface-type device exhibited higher suitability for the neuromorphic, symmetric, and linear long-term potentiation (LTP) and long-term depression (LTD). These findings suggest better types of memristors for implementing ionic memristor-based ANNs among the two types of RS mechanisms.

Neural networks trained by backpropagation have achieved tremendous successes on numerous intelligent tasks. However, naïve gradient-based training and updating methods on memristors impede applications due to intrinsic material properties. Here, we built a 39 nm 1 Gb phase change memory (PCM) memristor array and quantified the unique resistance drift effect. On this basis, spontaneous sparse learning (SSL) scheme that leverages the resistance drift to improve PCM-based memristor network training is developed. During training, SSL regards the drift effect as spontaneous consistency-based distillation process that reinforces the array weights at the high-resistance state continuously unless the gradient-based method switches them to low resistance. Experiments show that the SSL not only helps the convergence of network with better performance and sparsity controllability without additional computation in handwritten digit classification. This work promotes the learning algorithms with the intrinsic properties of memristor devices, opening a new direction for development of neuromorphic computing chips.

Phase-change memory (PCM), a non-volatile memory technology, is considered the most promising candidate for storage class memory and neuro-inspired devices. It is generally fabricated based on GeTe–Sb2Te3 pseudo-binary alloys. However, natively, it has technical limitations, such as noise and drift in electrical resistance and high current in operation for real-world device applications. Recently, heterogeneously structured PCMs (HET-PCMs), where phase-change materials are hetero-assembled with functional (barrier) materials in a memory cell, have shown a dramatic enhancement in device performance by reducing such inherent limitations. In this Perspective, we introduce recent developments in HET-PCMs and relevant mechanisms of operation in comparison with those of conventional alloy-type PCMs. We also highlight corresponding device enhancements, particularly their thermal stability, endurance, RESET current density, SET speed, and resistance drift. Last, we provide an outlook on promising research directions for HET-PCMs including PCM-based neuromorphic computing.

No abstract