Batching for Green AI -- An Exploratory Study on Inference

TL;DR

This study explores how input batching during inference impacts energy consumption and response times in neural networks, revealing significant effects and questioning the rapid rise in energy use relative to accuracy.

Contribution

It provides an empirical analysis of batching effects on energy and response times during inference and discusses the evolution of neural network efficiency over the last decade.

Findings

Batching significantly reduces energy consumption and response times during inference.

Energy consumption has increased faster than accuracy over the past decade.

ShuffleNetV2 offers a good balance of performance and low energy use.

Abstract

The batch size is an essential parameter to tune during the development of new neural networks. Amongst other quality indicators, it has a large degree of influence on the model's accuracy, generalisability, training times and parallelisability. This fact is generally known and commonly studied. However, during the application phase of a deep learning model, when the model is utilised by an end-user for inference, we find that there is a disregard for the potential benefits of introducing a batch size. In this study, we examine the effect of input batching on the energy consumption and response times of five fully-trained neural networks for computer vision that were considered state-of-the-art at the time of their publication. The results suggest that batching has a significant effect on both of these metrics. Furthermore, we present a timeline of the energy efficiency and accuracy of neural networks over the past decade. We find that in general, energy consumption rises at a much steeper pace than accuracy and question the necessity of this evolution. Additionally, we highlight one particular network, ShuffleNetV2(2018), that achieved a competitive performance for its time while maintaining a much lower energy consumption. Nevertheless, we highlight that the results are model dependent.