Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Survey of Sustainability in Large Language Models: Applications, Economics, and Challenges

Aditi Singh, Nirmal Prakashbhai Patel, Abul Ehtesham, Saket Kumar, Tala Talaei Khoei

TL;DR

The paper addresses the sustainability challenges posed by Large Language Models (LLMs), highlighting environmental, economic, and social considerations. It adopts a comprehensive survey approach, cataloging LLM applications, dissecting development and deployment costs, and detailing sustainable practices and lifecycle assessments, including a Be.Ta Labs case study. Its contributions include an application taxonomy, a cost framework, and actionable lifecycle guidance aimed at reducing energy consumption, emissions, and resource use. The findings are intended to guide researchers, industry practitioners, and policymakers toward renewable-powered, modular, and energy-efficient AI deployments that balance performance with environmental responsibility.

Abstract

Large Language Models (LLMs) have transformed numerous domains by providing advanced capabilities in natural language understanding, generation, and reasoning. Despite their groundbreaking applications across industries such as research, healthcare, and creative media, their rapid adoption raises critical concerns regarding sustainability. This survey paper comprehensively examines the environmental, economic, and computational challenges associated with LLMs, focusing on energy consumption, carbon emissions, and resource utilization in data centers. By synthesizing insights from existing literature, this work explores strategies such as resource-efficient training, sustainable deployment practices, and lifecycle assessments to mitigate the environmental impacts of LLMs. Key areas of emphasis include energy optimization, renewable energy integration, and balancing performance with sustainability. The findings aim to guide researchers, practitioners, and policymakers in developing actionable strategies for sustainable AI systems, fostering a responsible and environmentally conscious future for artificial intelligence.

A Survey of Sustainability in Large Language Models: Applications, Economics, and Challenges

TL;DR

The paper addresses the sustainability challenges posed by Large Language Models (LLMs), highlighting environmental, economic, and social considerations. It adopts a comprehensive survey approach, cataloging LLM applications, dissecting development and deployment costs, and detailing sustainable practices and lifecycle assessments, including a Be.Ta Labs case study. Its contributions include an application taxonomy, a cost framework, and actionable lifecycle guidance aimed at reducing energy consumption, emissions, and resource use. The findings are intended to guide researchers, industry practitioners, and policymakers toward renewable-powered, modular, and energy-efficient AI deployments that balance performance with environmental responsibility.

Abstract

Large Language Models (LLMs) have transformed numerous domains by providing advanced capabilities in natural language understanding, generation, and reasoning. Despite their groundbreaking applications across industries such as research, healthcare, and creative media, their rapid adoption raises critical concerns regarding sustainability. This survey paper comprehensively examines the environmental, economic, and computational challenges associated with LLMs, focusing on energy consumption, carbon emissions, and resource utilization in data centers. By synthesizing insights from existing literature, this work explores strategies such as resource-efficient training, sustainable deployment practices, and lifecycle assessments to mitigate the environmental impacts of LLMs. Key areas of emphasis include energy optimization, renewable energy integration, and balancing performance with sustainability. The findings aim to guide researchers, practitioners, and policymakers in developing actionable strategies for sustainable AI systems, fostering a responsible and environmentally conscious future for artificial intelligence.

Paper Structure

This paper contains 10 sections, 4 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Research Questions and Inquiries
  3. LLM Applications
  4. Cost of LLM Development
  5. Environmental Concerns
  6. Sustainable Green AI
  7. Sustainable Practices
  8. Lifecycle Assessments
  9. Case Study: Be.Ta Labs initiative towards sustainable AI
  10. Conclusion

Figures (4)

  • Figure 1: Hierarchical Tree of Research Questions and Objectives for LLM Sustainability
  • Figure 2: Carbon emissions for training BERT (8x V100s, 36 hrs): highest in Central US/Australia, lowest in Norway/France. From Dodge et al. dodge2022measuringcarbonintensityai
  • Figure 3: Carbon emissions for training experiments: larger models emit significantly more than smaller ones. From Dodge et al.dodge2022measuringcarbonintensityai
  • Figure 4: LLMCarbon Overview faiz2024