DeepEdgeBench: Benchmarking Deep Neural Networks on Edge Devices

TL;DR

This paper benchmarks the performance of various edge devices running deep neural networks, focusing on inference time and power consumption, and provides a measurement method for evaluating AI performance on resource-limited hardware.

Contribution

It introduces a comprehensive benchmarking framework for edge AI devices and compares multiple hardware platforms using different models and frameworks.

Findings

Google Coral Dev Board has the best performance for TensorFlow quantized models.

Nvidia Jetson Nano outperforms others for MobileNetV2 inference time under certain conditions.

The paper provides a flexible method for measuring power, inference time, and accuracy on edge devices.

Abstract

EdgeAI (Edge computing based Artificial Intelligence) has been most actively researched for the last few years to handle variety of massively distributed AI applications to meet up the strict latency requirements. Meanwhile, many companies have released edge devices with smaller form factors (low power consumption and limited resources) like the popular Raspberry Pi and Nvidia's Jetson Nano for acting as compute nodes at the edge computing environments. Although the edge devices are limited in terms of computing power and hardware resources, they are powered by accelerators to enhance their performance behavior. Therefore, it is interesting to see how AI-based Deep Neural Networks perform on such devices with limited resources. In this work, we present and compare the performance in terms of inference time and power consumption of the four Systems on a Chip (SoCs): Asus Tinker Edge R, Raspberry Pi 4, Google Coral Dev Board, Nvidia Jetson Nano, and one microcontroller: Arduino Nano 33 BLE, on different deep learning models and frameworks. We also provide a method for measuring power consumption, inference time and accuracy for the devices, which can be easily extended to other devices. Our results showcase that, for Tensorflow based quantized model, the Google Coral Dev Board delivers the best performance, both for inference time and power consumption. For a low fraction of inference computation time, i.e. less than 29.3% of the time for MobileNetV2, the Jetson Nano performs faster than the other devices.