Benchmarking Mobile Deep Learning Software
GetMobile: Mobile Computing and Communications · 22.10.2024 · 28 (3) · 5-8
DOI: 10.1145/3701701.3701703 ↗Kurz gefasst
Fachartikel in der Zeitschrift GetMobile: Mobile Computing and Communications, online erschienen am 22.10.2024. Verfasst von Qiyang Zhang und 1 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 12 Literaturangaben.
Abstract (Originalsprache)
Deploying deep learning (DL) on mobile devices has become increasingly prevalent. DL software libraries are crucial for efficient on-device inference, alongside algorithms and hardware. However, there has been limited understanding on the performance of modern DL libraries. We fill this gap by benchmarking 6 popular DL libraries and 15 diverse models across 10 mobile devices, which reveal an unsatisfactory landscape of mobile DL: their performance is highly disparate and fragmented across different models and hardware, and the impacts often surpass algorithm or hardware optimizations, such as model quantization and GPU/NPU-based computing. Finally, we provide practical implications for stakeholders in the DL library ecosystem, and envision a more ambitious picture of future mobile AI landscape in the LLM era.
Zitieren
APA
Zhang, Q., & Xu, M. (2024). Benchmarking Mobile Deep Learning Software. GetMobile: Mobile Computing and Communications, 28(3), 5-8. https://doi.org/10.1145/3701701.3701703
Vancouver
Zhang Q, Xu M. Benchmarking Mobile Deep Learning Software. GetMobile: Mobile Computing and Communications. 2024;28(3):5-8. doi:10.1145/3701701.3701703
BibTeX
@article{Zhang2024,
author = {Zhang, Qiyang and Xu, Mengwei},
title = {Benchmarking Mobile Deep Learning Software},
journal = {GetMobile: Mobile Computing and Communications},
year = {2024},
volume = {28},
number = {3},
pages = {5--8},
doi = {10.1145/3701701.3701703}
}
EUVIMED