AI Multi-Tenancy on Edge: Concurrent Deep Learning Model Executions and Dynamic Model Placements on Edge Devices

TL;DR

This paper explores AI multi-tenancy on edge devices, focusing on concurrent model execution and dynamic model placement, demonstrating significant throughput improvements and highlighting deployment opportunities for multiple deep learning services.

Contribution

It introduces and empirically evaluates techniques for AI multi-tenancy on edge devices, addressing a gap in prior research that focused mainly on single DL task performance.

Findings

Multi-tenancy improves inference throughput by up to 3.8x on Jetson TX2.

Techniques enable flexible deployment of multiple DL services.

Empirical evaluation across various devices and frameworks highlights benefits and limitations.

Abstract

Many real-world applications are widely adopting the edge computing paradigm due to its low latency and better privacy protection. With notable success in AI and deep learning (DL), edge devices and AI accelerators play a crucial role in deploying DL inference services at the edge of the Internet. While prior works quantified various edge devices' efficiency, most studies focused on the performance of edge devices with single DL tasks. Therefore, there is an urgent need to investigate AI multi-tenancy on edge devices, required by many advanced DL applications for edge computing. This work investigates two techniques - concurrent model executions and dynamic model placements - for AI multi-tenancy on edge devices. With image classification as an example scenario, we empirically evaluate AI multi-tenancy on various edge devices, AI accelerators, and DL frameworks to identify its benefits and limitations. Our results show that multi-tenancy significantly improves DL inference throughput by up to 3.3x -- 3.8x on Jetson TX2. These AI multi-tenancy techniques also open up new opportunities for flexible deployment of multiple DL services on edge devices and AI accelerators.