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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Embedding artificial intelligence at the edge (edge-AI) is an elegantsolution to tackle the power and latency issues in the rapidly expandingInternet of Things. As edge devices typically spend most of their time in sleepmode and only wake-up infrequently to collect and process sensor data,non-volatile in-memory computing (NVIMC) is a promising approach to design thenext generation of edge-AI devices. Recently, we proposed an NVIMC-basedneuromorphic accelerator using the phase change memories (PCMs), which we callas Raven. In this work, we demonstrate the ultra-low-power on-chip training andinference of speech commands using Raven. We showed that Raven can be trainedon-chip with power consumption as low as 30~uW, which is suitable for edgeapplications. Furthermore, we showed that at iso-accuracies, Raven needs 70.36xand 269.23x less number of computations to be performed than a deep neuralnetwork (DNN) during inference and training, respectively. Owing to such lowpower and computational requirements, Raven provides a promising pathwaytowards ultra-low-power training and inference at the edge.

Ultra-low power on-chip learning of speech commands with phase-change memories