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

Leveraging large data sets, deep Convolutional Neural Networks (CNNs) achievestate-of-the-art recognition accuracy. Due to the substantial compute andmemory operations, however, they require significant execution time. Themassive parallel computing capability of GPUs make them as one of the idealplatforms to accelerate CNNs and a number of GPU-based CNN libraries have beendeveloped. While existing works mainly focus on the computational efficiency ofCNNs, the memory efficiency of CNNs have been largely overlooked. Yet CNNs haveintricate data structures and their memory behavior can have significant impacton the performance. In this work, we study the memory efficiency of various CNNlayers and reveal the performance implication from both data layouts and memoryaccess patterns. Experiments show the universal effect of our proposedoptimizations on both single layers and various networks, with up to 27.9x fora single layer and up to 5.6x on the whole networks.

Optimizing Memory Efficiency for Deep Convolutional Neural Networks on GPUs