The global surge in the number of data centers is prompting major tech companies to explore sustainable energy solutions to support the growing demands of artificial intelligence (AI). Key strategies under consideration include a shift to nuclear energy, advanced liquid cooling techniques for data centers, and the integration of quantum computing.
Raj Hazra, CEO of Quantinuum, the largest integrated quantum computing firm, reflects on the cyclical nature of technology trends. “In the summer of every great technology, we discover there is a winter—don’t ignore it until it arrives,” he remarked, emphasizing the need to anticipate challenges as AI technologies expand.
As the demand for AI accelerates, critics urge that the tech industry must recognize the environmental costs associated with its rapid growth, moving away from the “move fast and break things” mentality. Somya Joshi, head of global agendas at the Stockholm Environment Institute, highlighted that the true environmental impact of AI remains obscured, largely because tech companies prioritize quick development and market entry.
According to the International Energy Agency, investments in data centers are projected to intensify, driven by increased digitalization and the rise of generative AI. This trend raises concerns about a significant spike in electricity demand, alongside the environmental implications of data center operations, which are essential for modern cloud computing and AI services.
Giampiero Frisio, president of electrification at ABB, noted that their data center business has experienced remarkable growth and is expected to rise by more than 24% in 2024. He emphasized that enhancing energy efficiency is paramount and that technology to achieve this, such as ABB’s HiPerGuard uninterruptible power supply (UPS), is readily available.
Frisio further explained the advantages of liquid cooling systems, which can effectively manage the increasing power density of modern data center equipment. “With the rise in computing power, the energy demands are expected to increase, so transitioning to liquid cooling is essential,” he stated, predicting that nuclear modular systems will play a significant role in the energy landscape in the next five to ten years.
In a notable trend, major U.S. tech firms like Microsoft, Google, and Amazon have recently secured multi-billion-dollar deals for nuclear energy, seeking to bolster their energy resources to support the computational needs of extensive AI models. This push for nuclear power coincides with a growing interest in liquid cooling methods, which utilize water to efficiently manage server temperatures. Schneider Electric recently made an $850 million acquisition of Motivair Corp, a leader in liquid cooling solutions, to enhance its offerings for data centers.
In addition to these traditional energy solutions, some tech leaders, including former Google CEO Eric Schmidt, have proposed that AI advancements could aid in decarbonizing data centers. However, Joshi dismissed this notion, arguing that relying on technology as a catch-all solution to environmental issues perpetuates a flawed narrative. She cautioned against assuming that exceeding planetary limits could be reconciled with technological fixes.
Hazra from Quantinuum articulated the importance of addressing sustainability within AI development. He pointed out that while the excitement surrounding generative AI is evident, there are two major concerns: resource sustainability and responsible usage. He believes that quantum computing could provide solutions to both challenges, allowing for more efficient and sustainable AI applications.
Hazra also noted a recent surge of interest from various businesses and investors in quantum technologies, highlighting a funding round that valued Quantinuum at $5 billion. He emphasized the necessity of presenting sustainable solutions to emerging problems rather than merely claiming to have answers.
“People will increasingly question the sustainability of their computing infrastructure,” Hazra predicted, envisioning a future where AI, quantum computing, and high-performance computing coexist to meet business needs while adhering to sustainability standards.
In conclusion, as Big Tech navigates the complex landscape of AI and energy consumption, the exploration of nuclear power, liquid cooling, and quantum computing could play pivotal roles in shaping a sustainable future for data centers and the technologies they support.
