English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Memristor-based neural networks provide an exceptional energy-efficientplatform for artificial intelligence (AI), presenting the possibility ofself-powered operation when paired with energy harvesters. However, mostmemristor-based networks rely on analog in-memory computing, necessitating astable and precise power supply, which is incompatible with the inherentlyunstable and unreliable energy harvesters. In this work, we fabricated a robustbinarized neural network comprising 32,768 memristors, powered by a miniaturewide-bandgap solar cell optimized for edge applications. Our circuit employs aresilient digital near-memory computing approach, featuring complementarilyprogrammed memristors and logic-in-sense-amplifier. This design eliminates theneed for compensation or calibration, operating effectively under diverseconditions. Under high illumination, the circuit achieves inference performancecomparable to that of a lab bench power supply. In low illumination scenarios,it remains functional with slightly reduced accuracy, seamlessly transitioningto an approximate computing mode. Through image classification neural networksimulations, we demonstrate that misclassified images under low illuminationare primarily difficult-to-classify cases. Our approach lays the groundwork forself-powered AI and the creation of intelligent sensors for variousapplications in health, safety, and environment monitoring.

Powering AI at the Edge: A Robust, Memristor-based Binarized Neural Network with Near-Memory Computing and Miniaturized Solar Cell