Generative AI on the Edge: Architecture and Performance Evaluation

TL;DR

This paper evaluates the performance of various Large Language Models on low-cost edge devices like Raspberry Pi, demonstrating their feasibility for localized AI inference in 6G networks without cloud dependence.

Contribution

It provides a systematic architecture and performance analysis of Generative AI models on commodity edge hardware, filling a gap in practical deployment insights.

Findings

Lightweight LLMs can run effectively on Raspberry Pi with acceptable throughput.

CPU-only deployment supports edge applications with low resource usage.

Edge GenAI enables localized inference in bandwidth-constrained environments.

Abstract

6G's AI native vision of embedding advance intelligence in the network while bringing it closer to the user requires a systematic evaluation of Generative AI (GenAI) models on edge devices. Rapidly emerging solutions based on Open RAN (ORAN) and Network-in-a-Box strongly advocate the use of low-cost, off-the-shelf components for simpler and efficient deployment, e.g., in provisioning rural connectivity. In this context, conceptual architecture, hardware testbeds and precise performance quantification of Large Language Models (LLMs) on off-the-shelf edge devices remains largely unexplored. This research investigates computationally demanding LLM inference on a single commodity Raspberry Pi serving as an edge testbed for ORAN. We investigate various LLMs, including small, medium and large models, on a Raspberry Pi 5 Cluster using a lightweight Kubernetes distribution (K3s) with modular prompting implementation. We study its feasibility and limitations by analyzing throughput, latency, accuracy and efficiency. Our findings indicate that CPU-only deployment of lightweight models, such as Yi, Phi, and Llama3, can effectively support edge applications, achieving a generation throughput of 5 to 12 tokens per second with less than 50\% CPU and RAM usage. We conclude that GenAI on the edge offers localized inference in remote or bandwidth-constrained environments in 6G networks without reliance on cloud infrastructure.