Memristive devices based hardware for unlabeled data processing

Abstract

Unlabeled data processing is of great significance for artificial intelligence (AI), since well-structured labelled data are scarce in a majority of practical applications due to the high cost of human annotation of labeling data. Therefore, automatous analysis of unlabeled datasets is important, and relevant algorithms for processing unlabeled data, such as k-means clustering, restricted Boltzmann machine and locally competitive algorithms etc., play a critical role in the development of AI techniques. Memristive devices offer potential for power and time efficient implementation of unlabeled data processing due to their unique properties in neuromorphic and in-memory computing. This review provides an overview of the design principles and applications of memristive devices for various unlabeled data processing and cognitive AI tasks.

Parallel in-memory wireless computing

“…Figure 1a shows the traditional digital technology and our parallel in-memory wireless computing scheme for parallel wireless digital communication. In-memory wireless computing fuses the signal processing and wireless transmission, and leverages the physical attributes Article https://doi.org/10.1038/s41928-023-00965-5 algorithms [33][34][35][36][37][38] and thus enable the direct processing of analogue signals to extract binary information without utilizing ADCs. Fusing in-memory computing based on the memristive crossbar with wireless digital communication in this way can realize parallel in-memory wireless computing and smoothen the sharp interconversion interfaces between the digital and analogue domains; it also has the potential to eventually eliminate the trade-off issues associated with DACs and ADCs.…”

Analysis, circuit design, and synchronization of a new 8D memristor-based chaotic system with application to chaotic masking and signal decoding

“…The simplicity of the design and compatibility with CMOS technology contribute to a considerable simplification of the memristor manufacturing process, thereby broadening its potential applications 3 . In addition to its compact size and low power consumption, the memristor possesses a distinctive feature -a memory function.…”

Editorial: Focus issue on energy-efficient neuromorphic devices, systems and algorithms

“…It also recommends reading in the RESET direction as preferable to the SET direction. The third paper in this series [9], authored by groups from Peking University, the Chinese Institute for Brain Research, and the Beijing Academy of Artificial Intelligence, focuses on using memristive devices for the processing of unlabeled data. This is a critical area in AI due to the scarcity of well-structured labeled data.…”