Search for: All records

Abstract Chalcogenide perovskites, particularly BaZrS₃, hold promise for optoelectronic devices owing to their exceptional light absorption and inherent stability. However, thin films obtained at lower processing temperatures typically result in small grain sizes and inferior transport properties. Here we introduce an approach employing co-sputtering elemental Ba and Zr targets followed by CS₂sulfurization, with a judiciously applied NaF capping layer. NaF acts as a flux agent during sulfurization, leading to marked increase in grain size and improved crystallinity. This process results in near-stoichiometric films with enhanced photoresponse. Terahertz spectroscopy further reveals a carrier mobility more than two orders of magnitude higher than those obtained from field-effect transistor measurements, suggesting that bulk transport is limited by grain boundary scattering. Our results demonstrate flux-assisted sulfurization as an effective strategy to improve the crystallinity of chalcogenide perovskite thin films for optoelectronic applications. Graphical abstract 

Spiking Neural Networks (SNNs) are energy-efficient artificial neural network models that can carry out data-intensive applications. Energy consumption, latency, and memory bottleneck are some of the major issues that arise in machine learning applications due to their data-demanding nature. Memristor-enabled Computing-In-Memory (CIM) architectures have been able to tackle the memory wall issue, eliminating the energy and time-consuming movement of data. In this work we develop a scalable CIM-based SNN architecture with our fabricated two-layer memristor crossbar array. In addition to having an enhanced heat dissipation capability, our memristor exhibits substantial enhancement of 10% to 66% in design area, power and latency compared to state-of-the-art memristors. This design incorporates an inter-spike interval (ISI) encoding scheme due to its high information density to convert the incoming input signals into spikes. Furthermore, we include a time-to-first-spike (TTFS) based output processing stage for its energy-efficiency to carry out the final classification. With the combination of ISI, CIM and TTFS, this network has a competitive inference speed of 2μs/image and can successfully classify handwritten digits with 2.9mW of power and 2.51pJ energy per spike. The proposed architecture with the ISI encoding scheme can achieve ∼10% higher accuracy than those of other encoding schemes in the MNIST dataset. 

The field of wireless communication is currently being pushed to new boundaries with the emergence of 6G technology. This advanced technology requires substantially increased data rates and processing speeds while simultaneously requiring energy-efficient solutions for real-world practicality. In this work, we apply a neuroscience-inspired machine learning model called echo state network (ESN) to the critical task of symbol detection in massive MIMO-OFDM systems, a key technology for 6G networks. Our work encompasses the design of a hardware-accelerated reservoir neuron architecture to speed up the ESN-based symbol detector. The design is then validated through a proof of concept on the Xilinx Virtex-7 FPGA board in real-world scenarios. The experiment results show the great performance and scalability of our symbol detector design across a range of MIMO configurations, compared with traditional MIMO symbol detection methods like linear minimum mean square error. Our findings also confirm the performance and feasibility of our entire system, reflected in low bit error rates, low resource utilization, and high throughput.