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

State-of-the-art systolic array-based accelerators adopt the traditionalim2col algorithm to accelerate the inference of convolutional layers. However,traditional im2col cannot efficiently support AI backpropagation.Backpropagation in convolutional layers involves performing transposedconvolution and dilated convolution, which usually introduces plenty ofzero-spaces into the feature map or kernel. The zero-space data reorganizationinterfere with the continuity of training and incur additional andnon-negligible overhead in terms of off- and on-chip storage, access andperformance. Since countermeasures for backpropagation are rarely proposed, wepropose BP-im2col, a novel im2col algorithm for AI backpropagation, andimplement it in RTL on a TPU-like accelerator. Experiments on TPU-likeaccelerator indicate that BP-im2col reduces the backpropagation runtime by34.9% on average, and reduces the bandwidth of off-chip memory and on-chipbuffers by at least 22.7% and 70.6% respectively, over a baseline acceleratoradopting the traditional im2col. It further reduces the additional storageoverhead in the backpropagation process by at least 74.78%.

BP-Im2col: Implicit Im2col Supporting AI Backpropagation on Systolic Arrays