A4 Vertaisarvioitu artikkeli konferenssijulkaisussa
Adaptive Workload Distribution for Accuracy-aware DNN Inference on Collaborative Edge Platforms
Tekijät: Taufique, Zain; Miele, Antonio; Liljeberg, Pasi; Kanduri, Anil
Toimittaja: N/A
Konferenssin vakiintunut nimi: Asia and South Pacific Design Automation Conference
Julkaisuvuosi: 2024
Journal: Proceedings of the Asia and South Pacific Design Automation Conference
Kokoomateoksen nimi: 2024 29th Asia and South Pacific Design Automation Conference (ASP-DAC)
Sarjan nimi: Proceedings of the Asia and South Pacific Design Automation Conference
Numero sarjassa: 29
Aloitussivu: 109
Lopetussivu: 114
ISBN: 979-8-3503-9355-2
eISBN: 979-8-3503-9354-5
ISSN: 2153-6961
eISSN: 2153-697X
DOI: https://doi.org/10.1109/ASP-DAC58780.2024.10473987
Verkko-osoite: https://ieeexplore.ieee.org/document/10473987
Preprintin osoite: https://arxiv.org/pdf/2310.10157
DNN inference can be accelerated by distributing the workload among a cluster of collaborative edge nodes. Heterogeneity among edge devices and accuracy-performance trade-offs of DNN models present a complex exploration space while catering to the inference performance requirements. In this work, we propose adaptive workload distribution for DNN inference, jointly considering node-level heterogeneity of edge devices, and application-specific accuracy and performance requirements. Our proposed approach combinatorially optimizes heterogeneity-aware workload partitioning and dynamic accuracy configuration of DNN models to ensure performance and accuracy guarantees. We tested our approach on an edge cluster of Odroid XU4, Raspberry Pi4, and Jetson Nano boards and achieved an average gain of 41.52\% in performance and 5.2\% in output accuracy as compared to state-of-the-art workload distribution strategies.