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

This work proposes a novel Energy-Aware Network Operator Search (ENOS)approach to address the energy-accuracy trade-offs of a deep neural network(DNN) accelerator. In recent years, novel inference operators have beenproposed to improve the computational efficiency of a DNN. Augmenting theoperators, their corresponding novel computing modes have also been explored.However, simplification of DNN operators invariably comes at the cost of loweraccuracy, especially on complex processing tasks. Our proposed ENOS frameworkallows an optimal layer-wise integration of inference operators and computingmodes to achieve the desired balance of energy and accuracy. The search in ENOSis formulated as a continuous optimization problem, solvable using typicalgradient descent methods, thereby scalable to larger DNNs with minimal increasein training cost. We characterize ENOS under two settings. In the firstsetting, for digital accelerators, we discuss ENOS on multiply-accumulate (MAC)cores that can be reconfigured to different operators. ENOS training methodswith single and bi-level optimization objectives are discussed and compared. Wealso discuss a sequential operator assignment strategy in ENOS that only learnsthe assignment for one layer in one training step, enabling greater flexibilityin converging towards the optimal operator allocations. Furthermore, followingBayesian principles, a sampling-based variational mode of ENOS is alsopresented. ENOS is characterized on popular DNNs ShuffleNet and SqueezeNet onCIFAR10 and CIFAR100.

ENOS: Energy-Aware Network Operator Search for Hybrid Digital and Compute-in-Memory DNN Accelerators