RESUMEN
Recent advancements have established quantum dots (QDs) as a promising alternative to conventional bulk materials in the fabrication of nanoscale integrated electronic devices. The appeal of QDs lies in their amenability to low-temperature processes and solution-based methodologies, facilitating the construction of devices with enhanced versatility and efficiency. The ternary metal chalcogenide CuInS2 QDs are one of the foremost, eco-friendly, and highly stable materials. In this study, CuInS2 QDs are employed as a functional layer in a memristive device featuring an Al/CuInS2/ITO configuration. The CuInS2 QDs have been synthesized by a hot injection method and characterized using X-ray diffraction (XRD) and transmission electron microscopy (TEM) to reveal their structural features. The Al/CuInS2/ITO device shows a unipolar resistive switching (RS) behaviour with a high on/off ratio of 105. The switching parameters have been studied for 100 cycles of SET/RESET. The SET and RESET voltages are found to be 1.66 ± 0.25 V and 0.69 ± 0.17 V. The spatial variability of switching parameters in the Al/CuInS2/ITO structure has also been studied for 9 different devices. The device also exhibits unipolar RS behaviour in the optimum temperature range of 0 °C to 50 °C. These outcomes demonstrate the impressive performance of CuInS2 QDs, indicating their potential for future energy-efficient and large-scale non-volatile memory applications.
RESUMEN
On-chip learning in spin orbit torque driven domain wall synapse based crossbar fully connected neural network (FCNN) has been shown to be extremely efficient in terms of speed and energy, when compared to training on a conventional computing unit or even on a crossbar FCNN based on other non-volatile memory devices. However there are issues with respect to scalability of the on-chip learning scheme in the domain wall synapse based FCNN. Unless the scheme is scalable, it will not be competitive with respect to training a neural network on a conventional computing unit for real applications. In this paper, we have proposed a modification in the standard gradient descent algorithm, used for training such FCNN, by including appropriate thresholding units. This leads to optimization of the synapse cell at each intersection of the crossbars and makes the system scalable. In order for the system to approximate a wide range of functions for data classification, hidden layers must be present and the backpropagation algorithm (extension of gradient descent algorithm for multi-layered FCNN) for training must be implemented on hardware. We have carried this out in this paper by employing an extra crossbar. Through a combination of micromagnetic simulations and SPICE circuit simulations, we hence show highly improved accuracy for domain wall syanpse based FCNN with a hidden layer compared to that without a hidden layer for different machine learning datasets.
