Sustainable AI and Green Computing: Reducing the Environmental Impact of Large-Scale Models with Energy-Efficient Techniques
Ojuawo Olutayo Oyewole, Jiboku Folahan Joseph
Abstract
Artificial intelligence (AI) has become an integral part of modern technology, driving advances across numerous sectors, including healthcare, finance, transportation, and entertainment. However, the rapid growth in AI model complexity particularly the rise of large language models has sparked concerns over their substantial energy consumption and associated carbon emissions. This paper explores the intersection of green computing and sustainable AI, focusing on the carbon footprint of large-scale models, energy-efficient algorithmic solutions, and emerging tools and frameworks designed to measure and mitigate environmental impact. We review current approaches such as model pruning, quantization, knowledge distillation, and efficient hardware, and discuss prominent tools like CodeCarbon and Carbontracker that enable researchers to track and reduce emissions. The paper also highlights ongoing challenges related to standardization, transparency, and policy, while outlining future research directions for creating an environmentally responsible AI ecosystem. By advancing sustainable AI practices, the research community can align innovation with environmental stewardship, ensuring that technological progress supports global climate goals.