Green LLM Techniques in Action: How Effective Are Existing Techniques for Improving the Energy Efficiency of LLM-Based Applications in Industry?

TL;DR

This study empirically evaluates various techniques to improve the energy efficiency of LLM-based industrial applications, finding that some methods significantly reduce energy use but often at the cost of accuracy, with only one method balancing both effectively.

Contribution

The paper provides the first empirical assessment of multiple energy-saving techniques applied to LLM applications in an industrial context, highlighting their effectiveness and limitations.

Findings

Prompt Optimization and 2-bit Quantization can reduce energy use by up to 90%.

Most techniques negatively impact accuracy, often beyond acceptable levels.

Small and Large Model Collaboration with NPCC achieves significant energy savings without major accuracy loss.

Abstract

The rapid adoption of large language models (LLMs) has raised concerns about their substantial energy consumption, especially when deployed at industry scale. While several techniques have been proposed to address this, limited empirical evidence exists regarding the effectiveness of applying them to LLM-based industry applications. To fill this gap, we analyzed a chatbot application in an industrial context at Schuberg Philis, a Dutch IT services company. We then selected four techniques, namely Small and Large Model Collaboration, Prompt Optimization, Quantization, and Batching, applied them to the application in eight variations, and then conducted experiments to study their impact on energy consumption, accuracy, and response time compared to the unoptimized baseline. Our results show that several techniques, such as Prompt Optimization and 2-bit Quantization, managed to reduce energy use significantly, sometimes by up to 90%. However, these techniques especially impacted accuracy negatively, to a degree that is not acceptable in practice. The only technique that achieved significant and strong energy reductions without harming the other qualities substantially was Small and Large Model Collaboration via Nvidia's Prompt Task and Complexity Classifier (NPCC) with prompt complexity thresholds. This highlights that reducing the energy consumption of LLM-based applications is not difficult in practice. However, improving their energy efficiency, i.e., reducing energy use without harming other qualities, remains challenging. Our study provides practical insights to move towards this goal.