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

In recent years, hardware-accelerated neural networks have gained significantattention for edge computing applications. Among various hardware options,crossbar arrays, offer a promising avenue for efficient storage andmanipulation of neural network weights. However, the transition from trainedfloating-point models to hardware-constrained analog architectures remains achallenge. In this work, we combine a quantization technique specificallydesigned for such architectures with a novel self-correcting mechanism. Byutilizing dual crossbar connections to represent both the positive and negativeparts of a single weight, we develop an algorithm to approximate a set ofmultiplicative weights. These weights, along with their differences, aim torepresent the original network's weights with minimal loss in performance. Weimplement the models using IBM's aihwkit and evaluate their efficacy over time.Our results demonstrate that, when paired with an on-chip pulse generator, ourself-correcting neural network performs comparably to those trained withanalog-aware algorithms.

Quantization of Deep Neural Networks to facilitate self-correction of weights on Phase Change Memory-based analog hardware