As data centre expansion, driven by artificial intelligence (AI) demand, continues, it is important to understand how those using data centres and those involved in building them align with the UN Sustainable Development Goals. Microsoft, Nvidia, and Taiwan Semiconductor Manufacturing Company (TSMC) are currently top 10 holdings within the EQ Positive Impact, Future Leaders & Climate Action Portfolios, and all are involved in the data centre value chain.
Nvidia offers a range of products including data centre systems, graphics processing units (GPUs), and networking equipment. Its CUDA software platform supports AI workloads, and its GPUs increasingly used in data centres, are essential for advanced computing tasks.
TSMC manufactures the semiconductor chips designed by Nvidia. These chips power a wide array of industries, from telecommunications to healthcare. TSMC’s role is crucial in translating Nvidia’s innovations into physical products.
We refer to the above companies as sustainability enablers, as they are companies that provide products and/or services that can be used by other companies to achieve their impact on social and environmental challenges.
Microsoft is an example of one of the end users of the products designed and manufactured by Nvidia and TSMC. Microsoft is one of the world’s leading computing providers and its products are used in many ways. Microsoft biggest revenue stream comes for its cloud computing product Microsoft Azure.
Cloud computing is considered a sustainable solution as it improves environmental outcomes for the end-user. Typically, cloud data centres use fewer servers and are powered more efficiently than traditional data centres, thus reducing the energy use and overall environmental impact for customers.
Data centres are resource intensive
Data centres and their energy usage have hit the headlines in recent weeks. It is estimated that global data centre electricity consumption currently accounts for 1-1.3% of global final electricity demand, with IEA projecting that electricity consumption for data centres could double by 2026.
Whilst there have been efficiency gains made, energy consumption from the likes of Amazon, Microsoft, Google, and Meta has more than doubled between 2017 and 2021. It was recently revealed that Alphabet’s (Google) carbon emissions have climbed by nearly 50% in the last 5 years due to AI demand. Its estimate that training AI models in data centres uses up 3 times more energy than traditional cloud tasks, and as the models continue to advance this will only increase the energy consumed. Water usage has also been a continued criticism of data centres. Water is used to cool down IT equipment and it’s been estimated that the average data centre uses a million litres of water a day which is roughly equivalent to the daily water use of 100,000 homes. These issues present the biggest challenges for data centres from an environmental standpoint.
How do the likes of Nvidia, TSMC and Microsoft mitigate these environmental challenges?
Nvidia and TSMC are actively working to mitigate the environmental impact of their chip innovation and production processes through various sustainability initiatives. Nvidia focuses on energy-efficient designs and targets net-zero greenhouse gas emissions by 2050. TSMC has implemented extensive water recycling and waste management programs within its manufacturing facilities. The company has also is on track to use 100% renewable energy by 2040 and significantly reduce its carbon footprint.
Importantly, both companies invest in research and development to create more energy-efficient chips, contributing to overall reductions in energy consumption and electronic waste in the AI value chain. Microsoft benefits from this, but also has committed to hourly match electricity use in its datacentres with renewable energy sources and has verified science-based targets. All three companies are considered ESG leaders (MSCI AA+).
What is the sustainability case for data centres?
The first point to make is that Nvidia has been at the forefront of innovation that has resulted in the world’s transition away from central processing units (CPUs) to the more energy efficient GPUs. Nvidia’s GPUs used to be sold for gaming end uses, but in the last two years, most of Nvidia’s revenue from the GPUs has come from data centre end uses (roughly 75% in 2024 so far).
When looking at the 500 most powerful data centres in the world, Nvidia products power 23 of the top 30 most energy efficient systems. Interestingly, when looking into the functions of the data centres that Nvidia’s products power, the majority are being used for academic or scientific research. Such insight on end-uses of the technology helps further build the sustainability case for the company.
Given the current level of transparency, we link Nvidia and TSMC to UN SDG 9, and more specifically to the sub target 9.5 “Enhance scientific research and upgrade the technological capabilities of industrial sectors in all countries”. We will continuously review the changing end-uses, alongside environmental externalities and efficiencies, when monitoring their UN SDG alignment.
Which other UN SDGs might be positively affected from using AI computing?
First, we should point out that no company is perfect. While Nvidia’s products do enable innovation across impactful industries, it also could be used for less impactful industries. With companies, such as Nvidia, that sit at earlier stage of a value chain it’s difficult to get full transparency on end uses. However, Nvidia has intentionally created products for certain industries that foster innovations with positive impacts on society and the environment.
Here are a few examples of how some of Nvidia’s products and services have been used.
Companies such as Trimble (currently held in the EQ Climate Action and Future Leaders Portfolios) use Nvidia’s products in its modelling to promote efficiency in industries such as agriculture. For instance, Trimble uses Nvidia’s GPUs to power its precision agriculture solutions., enhancing capabilities such as data processing and real time analytics.
Hubbell (currently held in the EQ Positive Impact and Climate Action Portfolios) provides hardware products used for by the power and utilities sector. They have partnered with Utilidata (software used for grid management), and Nvidia to help develop a module to help improve grid operations, with Hubbell integrating these modules into its hardware products. This collaboration helps in optimising energy distribution, reducing energy losses, and improving grid reliability.
Drug Discovery: biotech companies have used Nvidia’s products to accelerate drug discovery. For example, AstraZeneca (held in EQ Positive Impact, Future Leaders, and Climate Action Portfolios) has used Nvidia GPUs to accelerate the analysis of complex biological data, which is crucial for identifying potential drug candidates and understanding their interactions within the human body.
Improving Clinical trials: several companies are using Nvidia’s products to run AI models that help optimise clinical trials for the right patients, as the models are able to understand medical data nuances at scale, offering valuable real time insights. A notable example is Pfizer’s (held in EQ Positive Impact, Future Leaders, and Climate Action Portfolios) use of Nvidia GPUs to support its clinical trial data analysis in collaboration with the AI-driven platform from Medidata Solutions.
Improve digital surgery: companies have used Nvidia’s products to run AI models that give real-time surgical guidance, reducing surgical complication rates, thus improving patient outcomes. Intuitive Surgical (held in EQ Future Leaders, and Climate Action Portfolios), the company behind the da Vinci robotic surgical systems, utilises Nvidia products to enhance its digital surgery capabilities, particularly in improving the precision of its robotic systems. Specifically, Intuitive Surgical integrates Nvidia GPUs to accelerate real-time image processing and support advanced computer vision algorithms used in their surgical robots.
Summary
Without companies such as Nvidia, who have been pioneers in harnessing efficient computing power, innovation and future innovation across industries such as healthcare, agriculture, climate science and electric utilities might not be possible.
Whilst there are concerns about the electricity data centres consume in the short term, renewable energy matching and efficiency gains can mitigate this concern. It is likely that the potential long-term benefits from the innovation that comes from data centre usage is likely to improve the ability to mitigate and adapt to climate change, as well as drive other innovations that will contribute positively to other UN SDGs.