Managing the Environmental Impact of AI

Data centers are pivotal in powering a wide array of global technologies and services, ensuring crucial digital access, convenience, and efficiency in our daily, tech-based lives. Amazon's AWS data centers, for instance, are strategically located worldwide, providing reliable and scalable cloud services at the click of a button. However, data center construction, operation, and ongoing maintenance are having considerable environmental impacts that are attracting a lot of attention – and criticism – as their energy consumption grows. In fact, estimates suggest that data centers account for roughly 1-2% of the world's total electricity use, and according to a recent paper published by CellBy, “AI could account for anywhere between 0.3 to 0.5 percent of the world’s electricity use 4 years from now.” With AI-related activities consuming 10-20% of this energy use in the U.S. alone, projections indicate a major increase in the AI carbon footprint, with negative implications for future sustainability.

Defining Data Centers

A data center is a specialized facility designed to house computer systems and their associated components, such as telecommunications and storage. Data centers have evolved from small, mainframe rooms common in the 1940s to much larger modern facilities run by companies like Google, Microsoft, and Amazon that operate extensively worldwide to support cloud services and digital operations.

Often containing servers that provide computational and cyber defense systems, these centers are essential for ensuring data is accessible and secure while supporting critical business functions like email and e-commerce transactions. They also enable high-performance computing for data-intensive services such as AI, machine learning, and big data analytics.

Energy Use and AI Sustainability in Data Centers

Data centers facilitate incredible technological capabilities – but this comes at a cost. They are among the most energy-intensive building types, using 10-40 times more energy per square foot than commercial structures due to their need for a continuous and stable supply of electricity. From processing data to storing systems like hard drives and solid-state drives, powering networking equipment that manages data traffic, operating cooling systems to maintain optimal temperatures, and maintaining power systems for uninterrupted operation – servers require considerable energy to do their job.

Large-language models like ChatGPT, for example, can use anywhere from “85.4 to 134.0 TWh of electricity — the equivalent of Sweden or Argentina’s annual electricity use — by 2027,” which gives a worrisome glimpse into the future of power consumption when AI is involved. The International Energy Agency (IEA) predicts that global data center electricity demand will more than double from 2022 to 2026, largely due to the integration of AI into their operational systems. By 2030, AI could account for up to 5% of total U.S. power consumption, highlighting the need for innovative energy-efficient solutions and renewable energy adoption to address the environmental challenges of AI usage.

Controversy and Environmental Impact of Data Centers

Because of their high energy consumption, data centers are criticized for their environmental impact. They contribute roughly 0.3% of global carbon emissions – rising to 2% when including networked devices like laptops and smartphones – because the electricity used to power them often comes from fossil fuels and unrenewable sources. Data centers also require considerable amounts of water for cooling purposes, exacerbating water stress in regions already facing scarcity.

Additionally, the frequent upgrading and replacement of hardware components in data centers generate significant electronic waste that can contaminate soil and water if not disposed properly. The construction and operation of data centers demand large quantities of metals and other raw materials as well, and their physical installations and operations can lead to habitat destruction and pollution in local, rural areas.

Environmental Impact of AI

Despite these challenges, there are ongoing efforts to mitigate the environmental impact of data centers by combining technological innovation with sustainable practices – and AI is playing a crucial role as it might hold the key to its own sustainability. 

AI is being used to power predictive maintenance that analyzes sensor and equipment data to foresee component failures, allowing for timely maintenance that optimizes energy use and reduces downtime. AI is also optimizing systems through air-side economizers, liquid cooling, and cooling algorithms that adjust in real-time based on temperature and humidity data, significantly reducing the known temperature control issues of data centers. Google’s DeepMind AI continues to be trained to reduce energy used by this algorithmic method – and it was successful with a noted reduction at 40%. Nvidia has also recently made improvements towards sustainability by issuing its newest GPU hardware that will offer 25 times lower energy consumption than previous models.

On-site solar panels and wind turbines are becoming more common, and data centers are embracing sustainable practices by recycling and repurposing outdated hardware, thus extending the lifecycle of IT equipment and ensuring proper e-waste management. Creating data centers that do not require the use of a generator can also result in a cleaner, more efficient source of power.

These initiatives demonstrate the strides being made to reduce the environmental impact of data centers in moving toward an eco-friendly future – as “the growth in the industry shows no signs of slowing.”

Advancing Sustainable Technology Education

 

Capitol Technology University offers programs in the fields of study of Computer Science, Artificial Intelligence and Data Science, with degrees such as our Master of Research in Sustainability, that prepare students for innovative technology advancement with a focus on the ethical and environmental implications of these evolving industries. 

To learn more, contact our Admissions team or request more information.