Communication Dans Un Congrès Année : 2022

Co-Optimization of DNN and Hardware Configurations on Edge GPUs

Résumé

The ever-increasing complexity of both Deep Neural Networks (DNN) and hardware accelerators has made the cooptimization of these domains extremely complex. Previous works typically focus on optimizing DNNs given a fixed hardware configuration or optimizing a specific hardware architecture given a fixed DNN model. Recently, the importance of the joint exploration of the two spaces drew more and more attention. Our work targets the co-optimization of DNN and hardware configurations on edge GPU accelerators. We propose an evolutionary-based co-optimization strategy by considering three metrics: DNN accuracy, execution latency, and power consumption. By combining the two search spaces, a larger number of configurations can be explored in a short time interval. In addition, a better tradeoff between DNN accuracy and hardware efficiency can be obtained. Experimental results show that the co-optimization outperforms the optimization of DNN for fixed hardware configuration with up to 53% hardware efficiency gains with the same accuracy and inference time.

Fichier principal
Vignette du fichier
DSD22_Halima__Copy_.pdf (1.21 Mo) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Licence

Dates et versions

hal-03911415 , version 1 (22-12-2022)

Licence

Identifiants

Citer

Halima Bouzidi, Hamza Ouarnoughi, Smail Niar, El-Ghazali Talbi, Abdessamad Ait El Cadi. Co-Optimization of DNN and Hardware Configurations on Edge GPUs. 25th Euromicro Conference on Digital System Design (DSD), Aug 2022, Maspalomas, Gran Canaria, Spain. ⟨10.1109/DSD57027.2022.00060⟩. ⟨hal-03911415⟩
81 Consultations
333 Téléchargements

Altmetric

Partager

  • More